"... others have endorsed it explicitly and used it to argue that China’s economic model is superior"
Noah, this is actually not my position on this. My position is that China's economic model is fit to China and the unique set of conditions that it operates on at this point in history.
Broadly speaking, I do not believe you can wholesale remove a system and set of policies from their underlying conditions and apply it to another situation with a whole different set of underlying conditions.
Specifically with respect to industrial policy in the United States, I very much go back and forth on whether trying to implement China-style industrial policy in the United States would work (or even backfire) because of major differences in those underlying conditions.
What I do think is important is gaining an accurate understanding of how things work in China (including industrial policy) because there are direct impacts to the U.S. and global economy. Indeed, the optimal response may ultimately have to be doing something completely different because what works under one set of conditions (China) will often not work under a completely different set of ones (e.g. in the U.S.).
Just to add as well, the "lesson" I was referring to is very specifically about the need for industrial subsidies to be designed in a way that beneficiaries know they will be reduced over time in a relatively predictable manner.
This is to mitigate the risk that they view it as way to extract excess rent-seeking profits instead of actually investing and/or doing the hard work that helps the sector accomplish collective strategic goals, typically through learning spillovers.
If I am following (and pardon me if I am belaboring the obvious), the idea seems to be that a strong central government, such as in China, should be able to turn off the tap to subsidized companies that aren't making suitable progress.
But closer to home, with the divided US governmental system, the threat of ending subsidies may not be credible or reliable. Look how we struggle to close military bases; Congressfolks have lots of levers to pull to save a base or a plant in their district or state.
Or a more specific example: Iowa and corn subsidies for ethanol became a running quadrennial joke as Presidential aspirants for years would pander to Iowa farmers. Xi would never!
We can certainly debate the underlying causes but hard not to identify, as you have, how divisive politics can lead to less effective coordination and lowers effectiveness of subsidy programs.
That would at first glance seem to be an appropriate and correct response …. were it not for the fact that thousands of Chinese are fleeing from China and showing up at America’s southern border wishing to emigrate from China and start new lives in America. They do so for a variety of reasons: increased repression in China under Xi, steadily tightening political restrictions, unwinding of the freedoms many Chinese previously took for granted, the collapse of the Chinese stock market and real estate market. However, one of the main reasons cited is that China’s economy is failing; whether it is publicly reported in China or not, the fact is that unemployment is rising, businesses are failing, young people are having a hard time finding work. This has a huge impact on their life trajectories in terms of being able to afford to buy an apartment and/or get married. So people in China are effectively “voting with their feet”.
If the Chinese model was actually working as well as some like to believe, people would not be leaving to come to America.
People do not flee from a country where the economy is working well, where people are prospering, where there is hope for the future.
They flee because things are going so badly, they can see no future for themselves if they stay.
No, they are crossing the border illegally. Applying for asylum is legal, although as Noah outlined in a post I QT'd, this process is also being modified to close loopholes allowing non-authentic asylum seekers expedited legal processing.
This is faulty logic. By such logic, you can argue there is no such thing as illegal immigration because anyone can request asylum. Most of the asylum seekers are not real asylum seekers but seeking to exploit a loophole (that is now being closed).
I like this a lot. I’ve always opposed your support for industrial policy because I assumed it was rooted in arguments for economic vitality. But I would agree; national security is unambiguously a benefit of effectively targeted industrial policy.
I would also say the concern here is that virtually any rent seeking industry is going to make the national security argument because they know it resonates. And I think the US has really fallen victim to that, as its proponents in DC have also tried to make things about national security which are not about national security.
This argument is also interesting because it is disjointed from what appears to be the primary reason industrial policy enjoys whatever popular support it enjoys: 1. Bringing low skilled jobs back to the US (I would argue that industrial policy to achieve this is like putting a band aid on a gunshot wound; it’s a superficial solution that doesn’t address the underlying problem and 2. As a means of countering globalization.
A tiny point, but Han Feizi 韓非子 is also a person, not just a book--the zi 子 is an honorific, so Han Feizi means something like "Teacher Han Fei". I wonder why they chose him; the story is that Han Fei was betrayed by a friend and executed.
I wonder what rough estimates might be of what the dollar/yuan value would have been of consumer and producer surpluses without subsidies, and then what those values are with subsidies (and what the value of subsidies is)?
Perhaps much of the consumer surplus increase is an an inefficient subsidy pass through?
Maybe worth discussing in depth what "better" means in terms of running an economy and what metrics can be used to measure it. Are we going for pure GDP maxing here? This seems to be the lazy default of economics but isn't a very satisfying answer. e.g. increasing the work week from 40 to 60 hours will increase GDP by a lot, but I don't think many people would call it better.
Also I think it's hard to use China as a generalized example for anything in economics because it is still in (or has been until very recently) a unique rapid industrialization phase triggered by moving off of the stupidest economic system known to man.
In a free economy, the work week doesn't just change from 40-60 for no reason, it would change because consumer preferences prefer the money from those 20 hours to buy things rather than having them free. Obviously, there are many metrics we personally care more about in our lives, but they do not scale well to societies as a whole.
I can talk about life expectancy, morbidity, free time, or family formation, one day we will probably be able to measure percent of life experiencing joy or sadness, but how we value these many metrics against each other is different from person to person. I might want to live to 100 even if the last 10 years of my life are painful, full of chemo and surgeries, whereas someone else would rather only live to 92 but have those last 2 years not marred with painful medical procedures. Or I might value having enough money to raise a child at that additional 20 hours a week, but someone else would rather be child-free and have those 20 hours to themselves. The only way to aggregate all these preferences is GDP.
GDP growth and, to a lesser extent, GDP per capita and median income scale well under the assumption that price signals are mostly accurate representations of consumer preferences. It is absolutely important to look at those other metrics and try to get a more granular understanding. Still, it is also essential to have a profession dedicated to the big picture.
You're correct that GDP maxing isn't necessarily the goal, but as Noah has pointed out in the past and you can verify yourself countries with high GDP tend to correlate very well with countries that 1) show high levels of happiness, and 2) that I'd personally consider good places to live.
That certainly seems true overall but that also includes third world countries. For most countries at this time GDP maxing would be a solid strategy. But if you want to decide how to manage your economy once you get there, it's a less appealing strategy. Does the $76K GDP per capita in the US really translate into very much meaningful improvement over the $54K and $46K in Canada and the UK?
That's a difficult question, but my gut feeling is that having an extra $20,000 * 300 million goes a long way toward being able to address pressing problems like climate change.
It's difficult to compare North America with Europe, but comparing European countries, but having lived there, the standard of living in high GDP Switzerland is clearly higher than that of UK or neighboring France.
Can a country eg USA that has already substantially deindustrialised revert to re-industrialise the economy? Rebuilding clusters and supply chains is a lengthy process.
Most CEO's are not incentivized to think long term, in fact the opposite. Anything that has a cost that does not positively impact the bottom line (and their bonus) in a year or 2 generally doesn't even make it to the discussion stage. When moving from China they are going to other low cost countries and not to the USA. Noah has rightfully argued that building a collation in the world against China/Russia is in every democracy's best interests.
Yes and no-If they are a public traded company yes they have to deliver every qtr. Private companies taking a longer view because they can and its the most efficient method/process to run a company.
This is a question not a statement...could changes to how we tax capital gains and possibly things stock options impact this? I am not very knowledgeable in this area but I am wondering if the short-sightedness of publicly traded companies is inherent to the fact that they are publicly traded or more an artifact of how we tax the gains associated with them?
The metric we would want to influence is the median length an investor expects to hold in a given company. In the US, capital gains for assets held less than 1 year are roughly double that for assets held over a year. This helps stabilize financial markets for various reasons, but if we want CEOs to look to the long term, having lower rates for 5, 10, and 15 years would help. However, it would also have negative effects.
It would increase inequality; you are giving Elon Musk and Jef Bezos a much lower tax rate than an upper middle class person, who buys Apple for a few years and then sells it for mining or something. Specificaly its a transfer from the middle class and rich to the ultra rich.
It would reduce price discovery; the dominance of passive investing is already becoming a problem for markets, and this would effectively tax the main productive activity of financial markets, discovering the price of an asset by trading it. This could lead to larger and more destabilizing bubbles.
Is the rise of China likely to permanently discredit free trade as a belief system, causing countries to see industrial capacity (once built up) as something which ought to be defended to the utmost, even to the point of outright banning the foreign competition?
There's really no good reason why it can't. The US has 1) a highly skilled workforce, 2) incredible natural resources and 3) ample capital and existing stock (factories, etc).
Reasons to be doubtful do exist though. NIMBYism is a problem that isn't going to go away anytime soon. What we'll probably see is a huge transfer of wealth from existing rich states like CA, NY, MA to large low regulation states like Texas, Tennessee, Arizona. Note that this isn't my preference, but it appears to be inevitable unless something changes.
Also there is a serious problem with cost of labor. It is going to be hard to compete with countries that can pay workers 10% of what we can in the US does in any labor intensive manufacturing process.
One solution is to invest a lot in automation technology which can equalize this, but this is a 20 year plan and not nearly fast enough to counter China. In the short term I think our best bet is to move lower tech mass weapons manufacturing to Mexico and Central America.
The USA has not been a manufacturing based economy since the 80s and there is no going back, we are a service and technology based economy on paper. Here is something to consider that most outside the USA do not understand and its all 100% fact. 98% of all USA based businesses employ 100 or less people-small business is what really drives the USA economy and its not google and amazon!
China’s advantages are scale, forced tech sharing, and no permits or enviro concerns for connected projects. Takes ten years to build a big box store in American blue states. In China you can bulldoze a mountain and build a factory in months. I am not sure their policy of subsidizing particular products/indistries is going to work, but can’t fault them for going all in.
The Koreans and Japanese had an overall system that benefited oligarchic chaebols and trading companies (including protectionism and favorable credit terms). It wasn’t so much about individual products (though of course Japan has subsidized individual products/technologies). What is good for Samsung or Mitsubishi is good for the country (and unlike GM, they don’t mostly make one product).
The Dems problem is they wish to make the system as hostile as possible (regulations, taxes, union rules) but then hand out specific favors to help overcome these. Funny enough, this is how blue states run property development and permitting.
Rather than creating a system where affordable housing can be built, friendly developers are chosen to build “affordable” (ie subsidized, not naturally cheap) units as part of bigger projects that their donor recipients approve of.
The American (particularly the Dem) approach is incoherent and is more about payoffs than creating an ecosystem. Most of what Trump says or does is incoherent.
Industrial policy works well for developing countries because they have a road map provided by developed countries. Textiles > Steel > Autos > Electronics > etc. The things that make you good at following a map are a hinderance when you have to chart your own course.
Hi Noah, i think you are wrong on the point of entry subsidies - “This is a nuanced conclusion. It implies that China’s subsidies of computer chips, batteries, solar, steel, and other upstream goods are good for competition, but that subsidies of downstream products like EVs are not so great.”
Thats a wrong analogy to “mining, oil, forestry”. Right analogy is subsidising chemicals, germanium processing, silicon manufacturing etc. Chips and solar (no matter how much you like them), are already to far downstream, if we will follow this paper and their choice of industries to subsidise.
Well, the paper explicitly names semiconductors as the classic upstream industry that countries subsidize. And energy is as upstream a commodity as you can get!
Fair on semi although its midstream realistically at best. On energy, it would then be electricity, not solar panels. Just like one subsidises oil, not the rigs that get it out of the ground. Maybe its just a matter of classification and the lack of understanding of complexity of the industry by the authors. Upstream/downstream by itself is incomplete analysis. Example: fluoropolymer membrane -> electrolyser -> hydrogen. And thats a simple one, and here we already observe 3 levels. More broadly, i think its a good article and we need more of those. What you also fail to distinguish though is the question of motivation. You support industrial policy for national security reasons. Therefore, certain industries will be subsidised. Chinese use subsidies for multiple reasons and certainly dont subsidise everything (no kid toys get public money). So an interesting follow up could be on motivations for industrial policy vs industries chosen vs potential positives/negatives stemming from that decision. The other point is that you focus a lot on negatives as per economic theory (ie if industry x is subsidised it takes resources from industry y). While this might be academically correct, economic theory is south of politics, and policymakers might choose that this is the right way. Finally, you didnt address a fair criticism by the Chinese author/s - what the hell does stock market in US does anymore? Crazy overvaluations of a few stocks, undervaluations of many manufacturing stocks…what good does the current version of US stock market actually does for the US economy and how could it be reformed?
In traditional energy (e.g. oil) the classifications are:
- Upstream - Exploration & production
- Midstream - Pipelines & transport
- Downstream - Retail and marketing
Solar PV is different from an extractive industry like oil and so I don't think the same framework applies in an apples-to-apples comparable way.
Instead, I'd break it up into two segments: Solar PV mfg and Solar Grid Infrastructure with different process-driven value segmentation:
Solar PV mfg:
- Upstream: Refining and production of key commodity inputs (polysilicon refining, steel, aluminum, glass); specialized machinery and tools
- Downstream manufacturing steps: Wafer cutting --> Solar Cell manufacturing --> Solar module --> Deployment into the power grid (or off-grid applications)
Once installed in the grid, Solar PV end product simply becomes part of broad energy infrastructure, which itself is a primary input back every segment of the economy, including circularity with upstream solar PV mfg (particularly electricity-intensive polysilicon production)
Regarding national security, I think it's important to have a clear threat model in mind when considering this point and I'm not clear once you do that it favors any industrial policy beyond making sure China doesn't dominate chip production so much our military has to use potentially backdoored chips.
The threats the US faces are either long lasting smaller conflicts that leave us free to trade with other countries or true high intensity conflict with a near pear advisory like China which will be decided in too short a time frame to make manufacturing relevant. As such we would be better served by maximizing what our money can purchase to build up weapon stockpiles even if that means many of them get made in South Korea, Germany or Japan.
There just isn't really any likely threat that puts the US into the situation Russia finds themselves in -- rendered a pariah a blockaded for a long period so that we are unable to import materials or shells. Any attempt by China to blockade US ports of US shipments to/from South Korea would quickly escalate into a very intense hot war.
"which will be decided in too short a time frame to make manufacturing relevant"
This is what they say about every war and are wrong. e.g. US Civil War, WWI, WWII from Japan's perspective, Ukraine invasion from Russia's perspective.
If the US wins a short engagement in the South China Sea, do you really think China will just take the humiliation and move on when they have the option of continuing the war and pressing their industrial production advantage?
Maybe a better way to put the point is: if we don't insist on having weapons manufacturing plants in Alaska and Hawaii why do we need them inside the US rather than sourced from friendly countries? I don't think anyone believes the European countries all need to make their arms at home rather than in closely aligned EU countries.
We shouldn't be planning to pull a Russia and become a pariah. We should be thinking of ourselves as the key player in a collection of like minded countries and I'm sure that we can convince NATO countries as well as Japan, South Korea and Australia to sign agreements pledging to continue production in a conflict.
The problem NATO faces ramping up production is that they haven't been willing to make the long term commitments necessary to arms manufacturers needed to convince them to open up new plants. But that's a problem that's independent of where the weapons are produced and is fixed simply by making those deals.
If we buy our 155 artillery from South Korea and we need them to ramp up production in a conflict we do the same thing we would do for a company in the US -- promise to buy the full output of a new facility for the next 10 years.
Indeed, in a prolonged conflict where we might need to implement a draft again the fact that the shell production doesn't need to compete for workers with the military or other viral us industries is a plus.
I'm not claiming that there couldn't be any conflict that continued for a long period -- merely that such a conflict wouldn't cut us off from international trade with friendly nations. If you want to use history as a guide note that even in WW2 with the Nazi and japanese attacks on shipping we were still able to move huge amounts of weapons across the oceans
Also, I doubt China would continue fighting while it returned to reequip. I mean why would they continue to accept losses when they could just declare a ceasefire and rearm while they aren't under attack?
Besides, in such a conflict it's probably more convenient anyway to source a bunch of our weapons from various other countries anyway as it discourages direct attack on those plants.
--
But yes, obviously we shouldn't source our weapons directly from China but I took that pretty much for granted.
First, the billion Chinese are humans just like you and I and by buying from them we've hugely reduced global poverty while reducing many of our own costs.
Second, I still think the basic idea that a richer society is less inclined to go to war -- they have more to lose and invest more in children and are more aware of the world -- is correct. And they tend to have more pressure to be an open society. Being an autocratic society means that these effects aren't as strong directly but maintaining popular support still matters not to mention those social changes affect CCP leaders as well.
The idea this approach is discredited because it didn't happen overnight in China is crazy. It's a general trend not an overnight guarantee. Not a coincidence that NK is one of the poorest countries.
However unless a war with China went nuclear, the lesson of Ukraine is that it would probably NOT be over very quickly unless both sides decided they had achieved enough of their war aims to stop. We’ve forgotten that wars between peer adversaries don’t go to plan and do consume unbelievable amounts of materiel (as well as blood and treasure).
Disagree completely. Ukraine is going the way it is because of several factors.
1) The more powerful country (Russia) is trying to take and hold land occupied by an unfriendly force.
If the us needed to defend Taiwan it doesn't need to force troops into Beijing it needs to destroy enough Chinese ships, missles etc for them to give up on the attempt. If all Russia needed to accomplish was to render Ukraine effectively incapable of launching an oversea invasion of southern Russia we would be done by now. And no world where we start slogging into mainland Chinese territory that it stays conventional.
2) Russia **isn't** being effectively blockaded by any country. They remain free to send ships around the world which return filled with weapons and Ukraine is shipping in weapons too so it's not actually a case where my claim fails. The problem Russia has is almost no one is willing to sell them any -- hence my point about being a pariah.
3) Russia is fighting in its own backyard. Unless you think we are going to go to war with mexico or Canada that means we are projecting force overseas likely via our navy/air force -- what would we be fighting with if our aircraft pulled back like Russia's have? If our carrier force remains potent and capable for months or years that almost certainly means China won't have the capacity to interdict global shipping -- unlike in Ukraine it won't be right off their shore limiting drone effectiveness.
4) We have a fundamentally different military doctrine than Russia reliant on air superiority not masses of ground troops. If we establish air superiority we win any defensive war (eg Taiwan) and if we can't then our carriers are likely sunk and our nearby fighter wings destroyed. We go nuclear or go home.
5) We would go nuclear if there was a sustained and effective blockade of our ports because that means we'd be responding to the loss of large numbers of capital ships with huge casualties and haven't backed out.
I mean that out of something like 250 countries in the world the only 2 countries willing to sell them arms are Iran and North Korea.
Not all countries are participating in the sanctions regime and some countries are willing to sell civilian technologies that have military uses (China...tho hardly a vassal if anything Russia is in the subservient position there) but that's not the same.
Any conflict with China has a strong potential to go nuclear quickly. If they did win a conventional victory in the South China Sea, I'd have to think the US would start nuking transports or even parts of Taiwan itself if that were the only way to keep it out of China's hands.
Our strategy in Cold War 1.0 was that while we maintained a formidable military, we didn't try to match the Soviets at their core strength (ground power in Central Europe). Eisenhower correctly realized that if the Soviets really did break through the Fulda Gap we could just nuke them. And he choose a much cheaper nuclear deterrent over conscription and spending enough to match the Soviets tank for tank and artillery tube for artillery tube in Germany.
We should do the same in Cold War 2.0. Make some reasonable military preparations but the idea that we are going to match the industrial output of 1.4 billion people with an average IQ of 105 slanted towards STEM talent? Good luck. Ain't happening.
If nukes didn't exist then maybe we would be in a pickle, but the truth is it's just not possible to conquer another country if a nuclear power really doesn't want it to. Ukraine happened because nobody really cares about Ukraine and its situation is complicated. If we don't want China taking Taiwan, there is an obvious path to stopping it.
Meanwhile, all this talk of how we want to regime change China and we need to "keep them down" by trying to cut off their industries makes such a nightmare scenario more likely. I don't think the fate of the free world is improved at all by denying the Chinese the right to sell me a cheap EV, if anything the opposite is true.
“Meanwhile, all this talk of how we want to regime change China and we need to "keep them down" by trying to cut off their industries makes such a nightmare scenario more likely. I don't think the fate of the free world is improved at all by denying the Chinese the right to sell me a cheap EV, if anything the opposite is true.” 100%!
Noah where does Taiwan fit in with your analysis here? The country has reached top 20 in word in per capita GDP (PPP terms). Their use of industrial policy over decades (e.g. directed bank lending) to build out the world’s leading semiconductor and electronics manufacturing sectors seems prima facia central to their economic success. Would be great to get your take on Taiwan.
Many countries succeed while using industrial policy extensively. Korea has used it even more than Taiwan, and is just as successful! But that doesn't mean all industrial policies are good.
Less directed bank lending and more macro policies like stable/advantaged currency (weak TWD), infrastructure (stable, subsidized electricity), workforce development (electrical engineers).
Different from Korea and Japan, Taiwan’s industrial base was built on capital-light, nimble family-run SMBs. Even TSMC when it started was capital-light (tiny compared to IDMs like Intel and TI) and importing used SME that was generations behind cutting edge. It mainly built its capital base over time through profits and reinvested earnings.
I find it most puzzling that a socialist country requires capitalist tools to support an economy. Between the USSR and PRC, there ended up distribution problems, resource misallocation, and other common problems. It’s ironic that communism never spent enough time in the think tank to invent an entirely new economic system with its own novel approach to distribution and production. The “centralized” system appears extremely underdeveloped for the complexity of the country it’s trying to run. In fact, it still requires an export market to sustain itself. It’s like using an 8088 CPU to run a modern PC with Windows 11. It has become an amalgamation of parts rather than a true system based on an ideology.
XJP talks about a system with "Chinese characteristics," which just aren't very adaptable or resilient. It is not a good, nor accurate, reflection on Chinese potential.
One could argue that some tasks don't need as many workers? To borrow an example from Yudkowsky, what exactly do guys with PhD in physics do in the essentially zero-sum business of stock market?
I know they are, that's an actual empirical example! But I don't see much bad in industrial policy discouraging them to via actually sending them to something valuable.
Your arguments against #1 are surprising, given the realities of the past 30 years of US industrial policy. It might be wrong from an 'economic modeling' perspective, but actually seems quite wise from a 'people on the ground' perspective.
You could *very* easily argue that keeping the industrial base in the US by subsidizing factory production of cars would be one of the best ways to benefit large portions of the population left behind when that manufacturing was brought overseas. Precisely *because* growth in one part of the economy does not directly translate to on the ground welfare for people.
The notion that we can retrain factory workers into nurses is one of the very worst policy failures and ideas of the past 40 years - it literally has not worked at any meaningful scale.
Industry can't live forever, and pain is inevitable, but the pain caused by market dynamics vs. the pain caused by government control aren't comparable at all.
It's clear to me China has had some major strategic successes in its economic policy; part of it unfortunately relies on the continuing practice of economists ignoring the parts of policy that don't neatly fit into the models. That generates a lot of 'alpha' China has seen, and we shouldn't be afraid to steal something when it clearly works.
1. China's subsidies (the form of industrial policy that energizes this discussion of industrial policy) are almost trivial in absolute size - 1.73% of GDP according to CSIS (https://www.csis.org/analysis/red-ink-estimating-chinese-industrial-policy-spending-comparative-perspective). If you could transform an economy into a raging bull and achieve the largest reduction of poverty in history in a remarkably short time with unimaginative subsidies like cheap credit and buy domestic policies, etc. with that level of support, why isn't everyone doing it? More to the point, why isn't every country rich already??????
2. Since the subsidies are usually described as "pervasive", they aren't all that deep for any sector. For example, the BYD subsidy is estimated at 3.6% of revenues (https://www.ifw-kiel.de/publications/news/chinas-massive-subsidies-for-green-technologies/). Huge? Hardly - BYD's profit margin on European sales fully absorbs the 17.5% countervailing duty and leaves lots of margin to spare. Detroit is dying behind larger tariff walls.
3. China's subsidies are also small relative to its R&D share of GDP of about 2.1% of GDP - given China's singular focus on technology, a significant portion of that subsidy flow can be seen as offsetting R&D outlays (money is fungible after all) and is thus an efficient development tool. To the extent this is true - and it certainly is true for semiconductors where China has been cut off from leading tech and must innovate on a forced march pace - this is not a China-specific "model" - it's garden-variety industrial development policy. And its working: China is joining the Global North as a knowledge-based economy - see my paper on the BRICS - https://papers.ssrn.com/abstract=4492261 or in abridged from through CIGI here: https://www.cigionline.org/articles/expansion-aside-the-brics-path-to-growth-is-unchanged-its-technology/
3. China's tax share of GDP is about 20% compared to an OECD average of 34%. So China uses SOEs to do a lot of what OECD governments do in support of business with "horizontal" industrial policy with tax and spend. China also accordingly uses state-owned banks to use credit to do what it can't do with tax revenues. I simply underscore that 1.73% of GDP is a drop in the bucket compared to the 14% of GDP difference in tax share. Note: there used to be a running joke in China about "one family, two systems", where one partner works in the private sector for the income and the other in an SOE for the benefits. That's probably mostly dissipated but probably not entirely and likely is a factor in the below-average profitability of SOEs since they are delivering societal functions to some extent.
4. Insofar as the subsidies flow through to export prices, they also boost China's imports - this is Lerner Symmetry - which Doug Irwin has characterized as the most important theory for trade policy (equivalent to comparative advantage for trade economics). Insofar as China's subsidies are pervasive, they mainly serve to expand China's trade share of GDP (imports as well as exports); insofar as they are uneven, they affect the pattern of trade, but significant impacts are generally countervailed by much larger CVD duties. See my paper China’s Subsidies Reconsidered: Implications of Lerner Symmetry and Ricardian Comparative Advantage - https://papers.ssrn.com/abstract=3789391
5. And finally, the idea that China's support for investment is welfare-reducing as evidenced by the low consumption share of GDP is mostly off-base: China's investment is all at market prices; China's consumption continues to be significantly outside the market as evidenced by the fact that China's economy at market prices is only about 53% of its size at purchasing power parity. Market-priced investment into the PPP-priced economy is about 22% of GDP, which is not excessive at a time of rapid technological change. I first pointed this out in 2004 when the PPP-market gap was much larger - the point remains valid today. See: China’s Post-WTO Accession Boom: Is China Overheating? - https://papers.ssrn.com/abstract=1556028.
Long story short, China developed because they invested heavily in their knowledge infrastructure (universities, patent examiners, US-style VCs and NASDAQ-type exchange, etc) and became a knowledge-based economy that is the fastest-growing recipient of international intellectual property payments. That's why their path to the technology frontier (including in semis) is open and why the tech containment policy is failing.
"... others have endorsed it explicitly and used it to argue that China’s economic model is superior"
Noah, this is actually not my position on this. My position is that China's economic model is fit to China and the unique set of conditions that it operates on at this point in history.
Broadly speaking, I do not believe you can wholesale remove a system and set of policies from their underlying conditions and apply it to another situation with a whole different set of underlying conditions.
Specifically with respect to industrial policy in the United States, I very much go back and forth on whether trying to implement China-style industrial policy in the United States would work (or even backfire) because of major differences in those underlying conditions.
What I do think is important is gaining an accurate understanding of how things work in China (including industrial policy) because there are direct impacts to the U.S. and global economy. Indeed, the optimal response may ultimately have to be doing something completely different because what works under one set of conditions (China) will often not work under a completely different set of ones (e.g. in the U.S.).
Edit - my thoughts in response: https://x.com/glennluk/status/1814356615476236792?s=61
Fair enough!
Appreciate the clarification in the text.
Just to add as well, the "lesson" I was referring to is very specifically about the need for industrial subsidies to be designed in a way that beneficiaries know they will be reduced over time in a relatively predictable manner.
This is to mitigate the risk that they view it as way to extract excess rent-seeking profits instead of actually investing and/or doing the hard work that helps the sector accomplish collective strategic goals, typically through learning spillovers.
Fascinating, thanks.
If I am following (and pardon me if I am belaboring the obvious), the idea seems to be that a strong central government, such as in China, should be able to turn off the tap to subsidized companies that aren't making suitable progress.
But closer to home, with the divided US governmental system, the threat of ending subsidies may not be credible or reliable. Look how we struggle to close military bases; Congressfolks have lots of levers to pull to save a base or a plant in their district or state.
Or a more specific example: Iowa and corn subsidies for ethanol became a running quadrennial joke as Presidential aspirants for years would pander to Iowa farmers. Xi would never!
We can certainly debate the underlying causes but hard not to identify, as you have, how divisive politics can lead to less effective coordination and lowers effectiveness of subsidy programs.
That would at first glance seem to be an appropriate and correct response …. were it not for the fact that thousands of Chinese are fleeing from China and showing up at America’s southern border wishing to emigrate from China and start new lives in America. They do so for a variety of reasons: increased repression in China under Xi, steadily tightening political restrictions, unwinding of the freedoms many Chinese previously took for granted, the collapse of the Chinese stock market and real estate market. However, one of the main reasons cited is that China’s economy is failing; whether it is publicly reported in China or not, the fact is that unemployment is rising, businesses are failing, young people are having a hard time finding work. This has a huge impact on their life trajectories in terms of being able to afford to buy an apartment and/or get married. So people in China are effectively “voting with their feet”.
If the Chinese model was actually working as well as some like to believe, people would not be leaving to come to America.
People do not flee from a country where the economy is working well, where people are prospering, where there is hope for the future.
They flee because things are going so badly, they can see no future for themselves if they stay.
There are more identifiable and specific reasons about illegal immigration of Chinese through the Darien Gap
https://x.com/GlennLuk/status/1798034813821304979
https://x.com/GlennLuk/status/1814794448506683851
There’s a much more thoughtful and credible explanation here ———->
*HINT: If China’s industrial plans were really working out so well… people would not be leaving.
Also, the people coming to the border are NOT “illegals” at all.
They are PERFECTLY LEGAL.
It is completely legal to come to the border and request asylum.
Why Chinese Cross U.S. Southern Border In Growing Numbers
https://www.csmonitor.com/USA/2024/0521/chinese-migrants-why-border-crossing
No, they are crossing the border illegally. Applying for asylum is legal, although as Noah outlined in a post I QT'd, this process is also being modified to close loopholes allowing non-authentic asylum seekers expedited legal processing.
They are not crossing illegally.
There is literally no such thing as “illegal entry” for asylum seekers.
It doesn’t matter how or where they entered the US.
They can still request asylum.
This is faulty logic. By such logic, you can argue there is no such thing as illegal immigration because anyone can request asylum. Most of the asylum seekers are not real asylum seekers but seeking to exploit a loophole (that is now being closed).
I like this a lot. I’ve always opposed your support for industrial policy because I assumed it was rooted in arguments for economic vitality. But I would agree; national security is unambiguously a benefit of effectively targeted industrial policy.
I would also say the concern here is that virtually any rent seeking industry is going to make the national security argument because they know it resonates. And I think the US has really fallen victim to that, as its proponents in DC have also tried to make things about national security which are not about national security.
This argument is also interesting because it is disjointed from what appears to be the primary reason industrial policy enjoys whatever popular support it enjoys: 1. Bringing low skilled jobs back to the US (I would argue that industrial policy to achieve this is like putting a band aid on a gunshot wound; it’s a superficial solution that doesn’t address the underlying problem and 2. As a means of countering globalization.
a like from the big man himself. i'll take the friday win.
Is this Morse code?
A tiny point, but Han Feizi 韓非子 is also a person, not just a book--the zi 子 is an honorific, so Han Feizi means something like "Teacher Han Fei". I wonder why they chose him; the story is that Han Fei was betrayed by a friend and executed.
Lovely piece.
I wonder what rough estimates might be of what the dollar/yuan value would have been of consumer and producer surpluses without subsidies, and then what those values are with subsidies (and what the value of subsidies is)?
Perhaps much of the consumer surplus increase is an an inefficient subsidy pass through?
Maybe worth discussing in depth what "better" means in terms of running an economy and what metrics can be used to measure it. Are we going for pure GDP maxing here? This seems to be the lazy default of economics but isn't a very satisfying answer. e.g. increasing the work week from 40 to 60 hours will increase GDP by a lot, but I don't think many people would call it better.
Also I think it's hard to use China as a generalized example for anything in economics because it is still in (or has been until very recently) a unique rapid industrialization phase triggered by moving off of the stupidest economic system known to man.
In a free economy, the work week doesn't just change from 40-60 for no reason, it would change because consumer preferences prefer the money from those 20 hours to buy things rather than having them free. Obviously, there are many metrics we personally care more about in our lives, but they do not scale well to societies as a whole.
I can talk about life expectancy, morbidity, free time, or family formation, one day we will probably be able to measure percent of life experiencing joy or sadness, but how we value these many metrics against each other is different from person to person. I might want to live to 100 even if the last 10 years of my life are painful, full of chemo and surgeries, whereas someone else would rather only live to 92 but have those last 2 years not marred with painful medical procedures. Or I might value having enough money to raise a child at that additional 20 hours a week, but someone else would rather be child-free and have those 20 hours to themselves. The only way to aggregate all these preferences is GDP.
GDP growth and, to a lesser extent, GDP per capita and median income scale well under the assumption that price signals are mostly accurate representations of consumer preferences. It is absolutely important to look at those other metrics and try to get a more granular understanding. Still, it is also essential to have a profession dedicated to the big picture.
You're correct that GDP maxing isn't necessarily the goal, but as Noah has pointed out in the past and you can verify yourself countries with high GDP tend to correlate very well with countries that 1) show high levels of happiness, and 2) that I'd personally consider good places to live.
That certainly seems true overall but that also includes third world countries. For most countries at this time GDP maxing would be a solid strategy. But if you want to decide how to manage your economy once you get there, it's a less appealing strategy. Does the $76K GDP per capita in the US really translate into very much meaningful improvement over the $54K and $46K in Canada and the UK?
That's a difficult question, but my gut feeling is that having an extra $20,000 * 300 million goes a long way toward being able to address pressing problems like climate change.
It's difficult to compare North America with Europe, but comparing European countries, but having lived there, the standard of living in high GDP Switzerland is clearly higher than that of UK or neighboring France.
Can a country eg USA that has already substantially deindustrialised revert to re-industrialise the economy? Rebuilding clusters and supply chains is a lengthy process.
Most CEO's are not incentivized to think long term, in fact the opposite. Anything that has a cost that does not positively impact the bottom line (and their bonus) in a year or 2 generally doesn't even make it to the discussion stage. When moving from China they are going to other low cost countries and not to the USA. Noah has rightfully argued that building a collation in the world against China/Russia is in every democracy's best interests.
Yes and no-If they are a public traded company yes they have to deliver every qtr. Private companies taking a longer view because they can and its the most efficient method/process to run a company.
This is a question not a statement...could changes to how we tax capital gains and possibly things stock options impact this? I am not very knowledgeable in this area but I am wondering if the short-sightedness of publicly traded companies is inherent to the fact that they are publicly traded or more an artifact of how we tax the gains associated with them?
The metric we would want to influence is the median length an investor expects to hold in a given company. In the US, capital gains for assets held less than 1 year are roughly double that for assets held over a year. This helps stabilize financial markets for various reasons, but if we want CEOs to look to the long term, having lower rates for 5, 10, and 15 years would help. However, it would also have negative effects.
It would increase inequality; you are giving Elon Musk and Jef Bezos a much lower tax rate than an upper middle class person, who buys Apple for a few years and then sells it for mining or something. Specificaly its a transfer from the middle class and rich to the ultra rich.
It would reduce price discovery; the dominance of passive investing is already becoming a problem for markets, and this would effectively tax the main productive activity of financial markets, discovering the price of an asset by trading it. This could lead to larger and more destabilizing bubbles.
Is the rise of China likely to permanently discredit free trade as a belief system, causing countries to see industrial capacity (once built up) as something which ought to be defended to the utmost, even to the point of outright banning the foreign competition?
There's really no good reason why it can't. The US has 1) a highly skilled workforce, 2) incredible natural resources and 3) ample capital and existing stock (factories, etc).
Reasons to be doubtful do exist though. NIMBYism is a problem that isn't going to go away anytime soon. What we'll probably see is a huge transfer of wealth from existing rich states like CA, NY, MA to large low regulation states like Texas, Tennessee, Arizona. Note that this isn't my preference, but it appears to be inevitable unless something changes.
Also there is a serious problem with cost of labor. It is going to be hard to compete with countries that can pay workers 10% of what we can in the US does in any labor intensive manufacturing process.
One solution is to invest a lot in automation technology which can equalize this, but this is a 20 year plan and not nearly fast enough to counter China. In the short term I think our best bet is to move lower tech mass weapons manufacturing to Mexico and Central America.
The USA has not been a manufacturing based economy since the 80s and there is no going back, we are a service and technology based economy on paper. Here is something to consider that most outside the USA do not understand and its all 100% fact. 98% of all USA based businesses employ 100 or less people-small business is what really drives the USA economy and its not google and amazon!
China’s advantages are scale, forced tech sharing, and no permits or enviro concerns for connected projects. Takes ten years to build a big box store in American blue states. In China you can bulldoze a mountain and build a factory in months. I am not sure their policy of subsidizing particular products/indistries is going to work, but can’t fault them for going all in.
The Koreans and Japanese had an overall system that benefited oligarchic chaebols and trading companies (including protectionism and favorable credit terms). It wasn’t so much about individual products (though of course Japan has subsidized individual products/technologies). What is good for Samsung or Mitsubishi is good for the country (and unlike GM, they don’t mostly make one product).
The Dems problem is they wish to make the system as hostile as possible (regulations, taxes, union rules) but then hand out specific favors to help overcome these. Funny enough, this is how blue states run property development and permitting.
Rather than creating a system where affordable housing can be built, friendly developers are chosen to build “affordable” (ie subsidized, not naturally cheap) units as part of bigger projects that their donor recipients approve of.
The American (particularly the Dem) approach is incoherent and is more about payoffs than creating an ecosystem. Most of what Trump says or does is incoherent.
Industrial policy works well for developing countries because they have a road map provided by developed countries. Textiles > Steel > Autos > Electronics > etc. The things that make you good at following a map are a hinderance when you have to chart your own course.
Hi Noah, i think you are wrong on the point of entry subsidies - “This is a nuanced conclusion. It implies that China’s subsidies of computer chips, batteries, solar, steel, and other upstream goods are good for competition, but that subsidies of downstream products like EVs are not so great.”
Thats a wrong analogy to “mining, oil, forestry”. Right analogy is subsidising chemicals, germanium processing, silicon manufacturing etc. Chips and solar (no matter how much you like them), are already to far downstream, if we will follow this paper and their choice of industries to subsidise.
Well, the paper explicitly names semiconductors as the classic upstream industry that countries subsidize. And energy is as upstream a commodity as you can get!
Fair on semi although its midstream realistically at best. On energy, it would then be electricity, not solar panels. Just like one subsidises oil, not the rigs that get it out of the ground. Maybe its just a matter of classification and the lack of understanding of complexity of the industry by the authors. Upstream/downstream by itself is incomplete analysis. Example: fluoropolymer membrane -> electrolyser -> hydrogen. And thats a simple one, and here we already observe 3 levels. More broadly, i think its a good article and we need more of those. What you also fail to distinguish though is the question of motivation. You support industrial policy for national security reasons. Therefore, certain industries will be subsidised. Chinese use subsidies for multiple reasons and certainly dont subsidise everything (no kid toys get public money). So an interesting follow up could be on motivations for industrial policy vs industries chosen vs potential positives/negatives stemming from that decision. The other point is that you focus a lot on negatives as per economic theory (ie if industry x is subsidised it takes resources from industry y). While this might be academically correct, economic theory is south of politics, and policymakers might choose that this is the right way. Finally, you didnt address a fair criticism by the Chinese author/s - what the hell does stock market in US does anymore? Crazy overvaluations of a few stocks, undervaluations of many manufacturing stocks…what good does the current version of US stock market actually does for the US economy and how could it be reformed?
Manufacturing solar panels is actually upstream of energy!
In traditional energy (e.g. oil) the classifications are:
- Upstream - Exploration & production
- Midstream - Pipelines & transport
- Downstream - Retail and marketing
Solar PV is different from an extractive industry like oil and so I don't think the same framework applies in an apples-to-apples comparable way.
Instead, I'd break it up into two segments: Solar PV mfg and Solar Grid Infrastructure with different process-driven value segmentation:
Solar PV mfg:
- Upstream: Refining and production of key commodity inputs (polysilicon refining, steel, aluminum, glass); specialized machinery and tools
- Downstream manufacturing steps: Wafer cutting --> Solar Cell manufacturing --> Solar module --> Deployment into the power grid (or off-grid applications)
Once installed in the grid, Solar PV end product simply becomes part of broad energy infrastructure, which itself is a primary input back every segment of the economy, including circularity with upstream solar PV mfg (particularly electricity-intensive polysilicon production)
Regarding national security, I think it's important to have a clear threat model in mind when considering this point and I'm not clear once you do that it favors any industrial policy beyond making sure China doesn't dominate chip production so much our military has to use potentially backdoored chips.
The threats the US faces are either long lasting smaller conflicts that leave us free to trade with other countries or true high intensity conflict with a near pear advisory like China which will be decided in too short a time frame to make manufacturing relevant. As such we would be better served by maximizing what our money can purchase to build up weapon stockpiles even if that means many of them get made in South Korea, Germany or Japan.
There just isn't really any likely threat that puts the US into the situation Russia finds themselves in -- rendered a pariah a blockaded for a long period so that we are unable to import materials or shells. Any attempt by China to blockade US ports of US shipments to/from South Korea would quickly escalate into a very intense hot war.
"which will be decided in too short a time frame to make manufacturing relevant"
This is what they say about every war and are wrong. e.g. US Civil War, WWI, WWII from Japan's perspective, Ukraine invasion from Russia's perspective.
If the US wins a short engagement in the South China Sea, do you really think China will just take the humiliation and move on when they have the option of continuing the war and pressing their industrial production advantage?
Maybe a better way to put the point is: if we don't insist on having weapons manufacturing plants in Alaska and Hawaii why do we need them inside the US rather than sourced from friendly countries? I don't think anyone believes the European countries all need to make their arms at home rather than in closely aligned EU countries.
We shouldn't be planning to pull a Russia and become a pariah. We should be thinking of ourselves as the key player in a collection of like minded countries and I'm sure that we can convince NATO countries as well as Japan, South Korea and Australia to sign agreements pledging to continue production in a conflict.
The problem NATO faces ramping up production is that they haven't been willing to make the long term commitments necessary to arms manufacturers needed to convince them to open up new plants. But that's a problem that's independent of where the weapons are produced and is fixed simply by making those deals.
If we buy our 155 artillery from South Korea and we need them to ramp up production in a conflict we do the same thing we would do for a company in the US -- promise to buy the full output of a new facility for the next 10 years.
Indeed, in a prolonged conflict where we might need to implement a draft again the fact that the shell production doesn't need to compete for workers with the military or other viral us industries is a plus.
I'm not claiming that there couldn't be any conflict that continued for a long period -- merely that such a conflict wouldn't cut us off from international trade with friendly nations. If you want to use history as a guide note that even in WW2 with the Nazi and japanese attacks on shipping we were still able to move huge amounts of weapons across the oceans
Also, I doubt China would continue fighting while it returned to reequip. I mean why would they continue to accept losses when they could just declare a ceasefire and rearm while they aren't under attack?
Besides, in such a conflict it's probably more convenient anyway to source a bunch of our weapons from various other countries anyway as it discourages direct attack on those plants.
--
But yes, obviously we shouldn't source our weapons directly from China but I took that pretty much for granted.
Shouldn't we try to import from little as possible from China (and other New Axis states) in all spheres, not just weapons?
Not at all.
First, the billion Chinese are humans just like you and I and by buying from them we've hugely reduced global poverty while reducing many of our own costs.
Second, I still think the basic idea that a richer society is less inclined to go to war -- they have more to lose and invest more in children and are more aware of the world -- is correct. And they tend to have more pressure to be an open society. Being an autocratic society means that these effects aren't as strong directly but maintaining popular support still matters not to mention those social changes affect CCP leaders as well.
The idea this approach is discredited because it didn't happen overnight in China is crazy. It's a general trend not an overnight guarantee. Not a coincidence that NK is one of the poorest countries.
Also the fact that we do trade so much makes it much less attractive to go to war for both of us which is a plus.
However unless a war with China went nuclear, the lesson of Ukraine is that it would probably NOT be over very quickly unless both sides decided they had achieved enough of their war aims to stop. We’ve forgotten that wars between peer adversaries don’t go to plan and do consume unbelievable amounts of materiel (as well as blood and treasure).
Disagree completely. Ukraine is going the way it is because of several factors.
1) The more powerful country (Russia) is trying to take and hold land occupied by an unfriendly force.
If the us needed to defend Taiwan it doesn't need to force troops into Beijing it needs to destroy enough Chinese ships, missles etc for them to give up on the attempt. If all Russia needed to accomplish was to render Ukraine effectively incapable of launching an oversea invasion of southern Russia we would be done by now. And no world where we start slogging into mainland Chinese territory that it stays conventional.
2) Russia **isn't** being effectively blockaded by any country. They remain free to send ships around the world which return filled with weapons and Ukraine is shipping in weapons too so it's not actually a case where my claim fails. The problem Russia has is almost no one is willing to sell them any -- hence my point about being a pariah.
3) Russia is fighting in its own backyard. Unless you think we are going to go to war with mexico or Canada that means we are projecting force overseas likely via our navy/air force -- what would we be fighting with if our aircraft pulled back like Russia's have? If our carrier force remains potent and capable for months or years that almost certainly means China won't have the capacity to interdict global shipping -- unlike in Ukraine it won't be right off their shore limiting drone effectiveness.
4) We have a fundamentally different military doctrine than Russia reliant on air superiority not masses of ground troops. If we establish air superiority we win any defensive war (eg Taiwan) and if we can't then our carriers are likely sunk and our nearby fighter wings destroyed. We go nuclear or go home.
5) We would go nuclear if there was a sustained and effective blockade of our ports because that means we'd be responding to the loss of large numbers of capital ships with huge casualties and haven't backed out.
I mean that out of something like 250 countries in the world the only 2 countries willing to sell them arms are Iran and North Korea.
Not all countries are participating in the sanctions regime and some countries are willing to sell civilian technologies that have military uses (China...tho hardly a vassal if anything Russia is in the subservient position there) but that's not the same.
As I commented at another Substack, "You can have a 72-hour war if you're set up for a 72-month one."
Right, hence the importance of huge stockpiles of weapons rather than huge factories.
Any conflict with China has a strong potential to go nuclear quickly. If they did win a conventional victory in the South China Sea, I'd have to think the US would start nuking transports or even parts of Taiwan itself if that were the only way to keep it out of China's hands.
Our strategy in Cold War 1.0 was that while we maintained a formidable military, we didn't try to match the Soviets at their core strength (ground power in Central Europe). Eisenhower correctly realized that if the Soviets really did break through the Fulda Gap we could just nuke them. And he choose a much cheaper nuclear deterrent over conscription and spending enough to match the Soviets tank for tank and artillery tube for artillery tube in Germany.
We should do the same in Cold War 2.0. Make some reasonable military preparations but the idea that we are going to match the industrial output of 1.4 billion people with an average IQ of 105 slanted towards STEM talent? Good luck. Ain't happening.
If nukes didn't exist then maybe we would be in a pickle, but the truth is it's just not possible to conquer another country if a nuclear power really doesn't want it to. Ukraine happened because nobody really cares about Ukraine and its situation is complicated. If we don't want China taking Taiwan, there is an obvious path to stopping it.
Meanwhile, all this talk of how we want to regime change China and we need to "keep them down" by trying to cut off their industries makes such a nightmare scenario more likely. I don't think the fate of the free world is improved at all by denying the Chinese the right to sell me a cheap EV, if anything the opposite is true.
“Meanwhile, all this talk of how we want to regime change China and we need to "keep them down" by trying to cut off their industries makes such a nightmare scenario more likely. I don't think the fate of the free world is improved at all by denying the Chinese the right to sell me a cheap EV, if anything the opposite is true.” 100%!
Noah where does Taiwan fit in with your analysis here? The country has reached top 20 in word in per capita GDP (PPP terms). Their use of industrial policy over decades (e.g. directed bank lending) to build out the world’s leading semiconductor and electronics manufacturing sectors seems prima facia central to their economic success. Would be great to get your take on Taiwan.
Many countries succeed while using industrial policy extensively. Korea has used it even more than Taiwan, and is just as successful! But that doesn't mean all industrial policies are good.
Yep, I'll definitely write more about Taiwan.
Less directed bank lending and more macro policies like stable/advantaged currency (weak TWD), infrastructure (stable, subsidized electricity), workforce development (electrical engineers).
Different from Korea and Japan, Taiwan’s industrial base was built on capital-light, nimble family-run SMBs. Even TSMC when it started was capital-light (tiny compared to IDMs like Intel and TI) and importing used SME that was generations behind cutting edge. It mainly built its capital base over time through profits and reinvested earnings.
"But what you don’t see on the graph..." I didn't know you had an internal Henry Hazlett, Noah. :-)
I find it most puzzling that a socialist country requires capitalist tools to support an economy. Between the USSR and PRC, there ended up distribution problems, resource misallocation, and other common problems. It’s ironic that communism never spent enough time in the think tank to invent an entirely new economic system with its own novel approach to distribution and production. The “centralized” system appears extremely underdeveloped for the complexity of the country it’s trying to run. In fact, it still requires an export market to sustain itself. It’s like using an 8088 CPU to run a modern PC with Windows 11. It has become an amalgamation of parts rather than a true system based on an ideology.
having said that, although the West is further along in understanding its own system, this is not to imply that it has been mastered to any extent.
XJP talks about a system with "Chinese characteristics," which just aren't very adaptable or resilient. It is not a good, nor accurate, reflection on Chinese potential.
> but that means fewer workers for other tasks.
One could argue that some tasks don't need as many workers? To borrow an example from Yudkowsky, what exactly do guys with PhD in physics do in the essentially zero-sum business of stock market?
You say that, but some Physics PhDs do get hired for quantitative finance. Quantitative finance and the stuff they learn are pretty related.
I know they are, that's an actual empirical example! But I don't see much bad in industrial policy discouraging them to via actually sending them to something valuable.
Ahh, thanks for clarifying your point.
Your arguments against #1 are surprising, given the realities of the past 30 years of US industrial policy. It might be wrong from an 'economic modeling' perspective, but actually seems quite wise from a 'people on the ground' perspective.
You could *very* easily argue that keeping the industrial base in the US by subsidizing factory production of cars would be one of the best ways to benefit large portions of the population left behind when that manufacturing was brought overseas. Precisely *because* growth in one part of the economy does not directly translate to on the ground welfare for people.
The notion that we can retrain factory workers into nurses is one of the very worst policy failures and ideas of the past 40 years - it literally has not worked at any meaningful scale.
Industry can't live forever, and pain is inevitable, but the pain caused by market dynamics vs. the pain caused by government control aren't comparable at all.
It's clear to me China has had some major strategic successes in its economic policy; part of it unfortunately relies on the continuing practice of economists ignoring the parts of policy that don't neatly fit into the models. That generates a lot of 'alpha' China has seen, and we shouldn't be afraid to steal something when it clearly works.
I would make several observations:
1. China's subsidies (the form of industrial policy that energizes this discussion of industrial policy) are almost trivial in absolute size - 1.73% of GDP according to CSIS (https://www.csis.org/analysis/red-ink-estimating-chinese-industrial-policy-spending-comparative-perspective). If you could transform an economy into a raging bull and achieve the largest reduction of poverty in history in a remarkably short time with unimaginative subsidies like cheap credit and buy domestic policies, etc. with that level of support, why isn't everyone doing it? More to the point, why isn't every country rich already??????
2. Since the subsidies are usually described as "pervasive", they aren't all that deep for any sector. For example, the BYD subsidy is estimated at 3.6% of revenues (https://www.ifw-kiel.de/publications/news/chinas-massive-subsidies-for-green-technologies/). Huge? Hardly - BYD's profit margin on European sales fully absorbs the 17.5% countervailing duty and leaves lots of margin to spare. Detroit is dying behind larger tariff walls.
3. China's subsidies are also small relative to its R&D share of GDP of about 2.1% of GDP - given China's singular focus on technology, a significant portion of that subsidy flow can be seen as offsetting R&D outlays (money is fungible after all) and is thus an efficient development tool. To the extent this is true - and it certainly is true for semiconductors where China has been cut off from leading tech and must innovate on a forced march pace - this is not a China-specific "model" - it's garden-variety industrial development policy. And its working: China is joining the Global North as a knowledge-based economy - see my paper on the BRICS - https://papers.ssrn.com/abstract=4492261 or in abridged from through CIGI here: https://www.cigionline.org/articles/expansion-aside-the-brics-path-to-growth-is-unchanged-its-technology/
3. China's tax share of GDP is about 20% compared to an OECD average of 34%. So China uses SOEs to do a lot of what OECD governments do in support of business with "horizontal" industrial policy with tax and spend. China also accordingly uses state-owned banks to use credit to do what it can't do with tax revenues. I simply underscore that 1.73% of GDP is a drop in the bucket compared to the 14% of GDP difference in tax share. Note: there used to be a running joke in China about "one family, two systems", where one partner works in the private sector for the income and the other in an SOE for the benefits. That's probably mostly dissipated but probably not entirely and likely is a factor in the below-average profitability of SOEs since they are delivering societal functions to some extent.
4. Insofar as the subsidies flow through to export prices, they also boost China's imports - this is Lerner Symmetry - which Doug Irwin has characterized as the most important theory for trade policy (equivalent to comparative advantage for trade economics). Insofar as China's subsidies are pervasive, they mainly serve to expand China's trade share of GDP (imports as well as exports); insofar as they are uneven, they affect the pattern of trade, but significant impacts are generally countervailed by much larger CVD duties. See my paper China’s Subsidies Reconsidered: Implications of Lerner Symmetry and Ricardian Comparative Advantage - https://papers.ssrn.com/abstract=3789391
5. And finally, the idea that China's support for investment is welfare-reducing as evidenced by the low consumption share of GDP is mostly off-base: China's investment is all at market prices; China's consumption continues to be significantly outside the market as evidenced by the fact that China's economy at market prices is only about 53% of its size at purchasing power parity. Market-priced investment into the PPP-priced economy is about 22% of GDP, which is not excessive at a time of rapid technological change. I first pointed this out in 2004 when the PPP-market gap was much larger - the point remains valid today. See: China’s Post-WTO Accession Boom: Is China Overheating? - https://papers.ssrn.com/abstract=1556028.
Long story short, China developed because they invested heavily in their knowledge infrastructure (universities, patent examiners, US-style VCs and NASDAQ-type exchange, etc) and became a knowledge-based economy that is the fastest-growing recipient of international intellectual property payments. That's why their path to the technology frontier (including in semis) is open and why the tech containment policy is failing.