The Dan Wang year-end letter

Always recommended.

This year I want to discuss mostly science and technology. First, some thoughts on China’s technology efforts. Then I’ll present a few reflections on science fiction, with a focus on Philip K. Dick and Liu Cixin. Next I’ll discuss books I read on American industrial history. I save personal reflections for the end.

Dan now lives in Beijing.  He left out music, however…


Good analysis, but like others in the #squad clearly too uncomfortable with people to be as influential as his writing seems to merit.

It's like this: failure is ok> that is the theme of Alice in Wonderland. A painter on that last brushstroke has failed. Even Van Gough who carried the burden! What a wonderful thought, that in a marriage ceremony equality is absent. Absent hope. And that's why I think we get in lost in this catastrophe….the closest thing to the American revolution is 1948....kaboom. Israel is created. Soviet Union, Iran, and the US are the first 3 to accept. And then for what its worth Herzl follows of all people john quincy adams into South America... And speaking of Russian, there is this thing! Its when a writer reads his own writing and its called when a lion kills. I killed bill. Take a sunset in Africa. Is that who he is? Herzl made one thing clear, with the clarity of salt! Pride is not evidence of belief. Perhaps Lebron James is the first person in the history of the world to realize this.

"And then for what its worth Herzl follows of all people john quincy adams into South America.."

That's pretty good.

Good analysis, though I find the one quantitative statistic the most interesting—China accounts for 20% of world manufacturing exports. This is unexceptional because China also has 20% of the world’s population. In many economic statistics, China tracks near the global average in per capita terms. This suggests that China is not a manufacturing powerhouse but is in many ways just the average and median country. In the long run, I’d expect it to contribute around 20% of global innovation.

How long is your “long run”? China has a lot of problems, but those are also opportunities. Right now China is aiming to get everyone in its next generation a high school education, and more people a college or post secondary education of some sort. Right now compulsory education only goes through 9th grade, and that means that if you don’t have good enough grades and test scores, the government will not allow you to go to a public high school. Some people can still afford to send their children to a private school, but my impression is that most people don’t have that kind of money so kids who don’t make the grade when they are 14 or 15 years old go immediately into full time, menial work. Also, a very large percentage of China’s population still lives in the countryside. I have read estimates that within 30 years or so, 70-80% of China’s population will live in cities, which means 200 million people or more moving to urban areas. So in twenty years the Chinese workforce will be much better educated, and a much greater proportion of it located in the most economically productive areas of China.

Another way to think about this is to think about “state capacity”. China has a good deal more state capacity than the global average, and the Chinese state is likely to become even more capable over the coming decades. So long as that capability is used judiciously, China should become much more productive, and hence it is reasonable to think that it should punch above the global average in terms of productivity and exports.

Does China have more state capacity than the global average though? It’s government plus state owned enterprise share of GDP is in the 20%s, similar to the US and much lower than Europe.

I agree that China still has opportunities and room to grow. Just think about how dysfunctional Greece is, yet Greece still has twice the GDP per capita of China. It seems a reasonable prediction that China could in the long run catch up to Greece. However, China would probably benefit from less state capacity at this point. One of the main things hindering urbanization is China’s own hukou laws that restrict movement. China’s state-owned enterprises are also less efficient and profitable than private ones.

I think of Mexico or India as having something close to global average state capacity. The world is mostly Africa and Asia, and Europe and North America are huge outliers in terms of state capacity.

Also, China has recently liberalized the hukou system. Residence cards for the tier one cities are still very hard to get, but my understanding is that the government wants people to live in cities, and so has changed the rules so that rural folks have at least some cities where they can easily get a residence card. Not ideal from a libertarian perspective by any means, but not nearly as bad as it once was.

In other words, engineering and innovation follows manufacturing....

Optimism may be politically correct for Chinese workers responding to surveys. What about incentives... aka freedom, rule of law, individual rights? If elites are so hedonistic, how do they survive? So all this stuff about elites being tied to family, marriage, education and hard work is nonsense? Perhaps we should distinguish between economic and political elites? The elites survived even in Heilein's dystopias. But everyone ignores HG Wells who probably knew more about science than anyone in his age and was closer to real developments than many of his peers yesterday or today. In epidemics, the human being comes back stronger every time. It takes faith. Climate change for past generations resulted in innovation, why not now? We should push ahead for alternatives to fossil fuel, not hide in a cave and wear plant sweaters fed with urine. How to neutralize poisonous plutonium? So if interchangeability is as important why do we have to "total" cars for insurance purposes? Is there a cost to inventory? The Egyptians had miniature steam engines, so why didn't they develop them? Didn't they have wars? Slavery maybe? So Britain ruled the US and that makes a sinister? What city burns the most coal? Some Chinese cities? You don't feel well? We need more animals and animal lovers? Really?

Rilke's panther, Monty Python, Trivers' lizard, Hendrix's zebra, the remains of hideousness; the clout of digestion, the idea of biological warfare. Yet there is a jaguar, yet there is cruelty, there is the kaliyanag, the raven, the vulture, the seraphim, the wildabeast; this jaguar is rare and beautiful but it is mutinous; a jaguar cares. a jaguar prowls and hides its eyes and takes some lever of pride. So what is woman to a god? It is and the more obvious it becomes the less you'll seek justice, hunting. It is an upturned grin. It is perdition, not purgatory. Prostitution not whoredom. The prostitute goes to the John.

It’s a shame we don’t hear from Dan more than once a year. Does he publish elsewhere any of his thought/writings? He is one of the more original voices out there. I keep checking his blog many times over the year but this letter seems to be it. A pity.

It seems that he has paying custom for his opinions now. One shouldn't really write for pleasure under those circumstances.

Some good stuff there. There were three items that I disagreed with or didn't understand.

"And so long as substantial US tariffs stay in place, Chinese firms will have worse access to the world’s largest and best consumer market, meaning that they’ll be exposed to less export discipline."

The first 2/3 of the sentence is fine, but I don't know what he means by "export discipline". And regardless of whatever meaning I attach to that phrase, I don't see how US tariffs on exports from China imply less discipline.

"Between its economic stagnation and general air of nostalgia, it’s difficult to identify anything in the streets of Hong Kong that didn’t already exist in the ‘90s."

I've never been to Hong Kong so I don't know exactly what he's referring to. But how much different is walking through any city nowadays compared to 20 years ago? Aside from some new buildings and businesses -- is that what he's referring to? Are Hong Kong's streets and skyline that much unchanged from 20 years ago?

"There are a few places that feel like the center of the world when you’re there, and Beijing is one of them. (I offer San Francisco, Tokyo, and DC as other candidates.)"

Hmmph. I'd put New York over SF and DC.

Plenty of good observations in the rest of the essay, although I think it would've benefited from being a good 1/3 or 1/4 shorter: the ratio of observations to words was pedestrian.

Export discipline is a polite way of saying fraud. When China was first entering global markets back in the 8os and 90s, textile imports were generally regulated by country specific quotas. The Chinese were adept at relabeling their textiles, and thus export more than their quota.

Expect countries like Vietnam or Malaysia to become major manufacturers of no-name manufactured households goods and electronics in a year or two if a truly effective tariff regime is in place.

Hong Kong is usually compared with mainland Chinese cities, which are all drastically different from 20 years ago. Hence the feeling of stagnation in Hong Kong. The reality is that Hong Kong should be compared with other developed cities in Japan, Europe and North America. The low hanging fruit has been picked, and Hong Kong is understandably no longer absorbing millions of migrants from the Chinese countryside.

So Hong Kong is stagnant/boring/etc. in the same way that New York, Tokyo, London, etc. are stagnant and boring.

I feel export discipline means: if you are selling to America, your stuff has to be as good as their domestic market leaders, for the price. If not, your stuff has to be as good as Chinese or European market leaders, which may well be a lower quality bar to pass. I feel this is the meaning because Wang seems a great believer in learning by doing and, more generally, in process thinking (c.f. state capacity libertarianism and the rest of the ideas of the #squad).

" the companies pull smart kids from R&D-intensive fields like materials science or semiconductor manufacturing, into ad optimization and game development."

This can't be emphasized enough. Unfortunately in the West, high finance also pulls talent away from these important fields. I suppose China's communist party also serves that function. Market is still the best system but it leaves gaps that could be exposed by anyone who can think beyond the next quarter.

Don’t intelligent and competent government officials create huge positive spillovers for a country? I thought that was a large part of Lee Kwan Yew’s successful plan to make Singapore a wealthy country; pay government officials very high salaries so that you get very intelligent and hard working people administering the state, and so that they have no need to take bribes to enhance their social status.

Why is this a problem? This just shows that ad optimization and game development delivers more benefits to consumers than more advances in material science and semiconductors. This makes sense to me as a consumer; today’s computer hardware is far more advanced than its software. I’d certainly get more utility out of more and better games and apps than faster computers and smartphones at this point, and this is even more true in emerging hardware fields like VR. This is a good example of the free market delivering what consumers rather than central planners want.

Think of finance, especially high finance if I understand you correctly, as being the capitalist system's investment allocation mechanism. Then finance looks hella important, maybe the most important thing.

"The legislators might in fact understand that semiconductors are a core strategic asset, in a way that social networks and search engines are not."

Social networks have massive blackmail potential and have fully realized disinformation campaigns being run on them. They are absolutely a core strategic asset that is why CFIUS cracked down on China trying to buy a few networks and the US is right to ask Facebook and Twitter to ban foreign psyops and election interference. The potential of your own government falling under the sway of a foreign power is too big a risk.

On the other point, if semiconductor fabrication is so important to national security then why did the US move the industry, which it invented, to Asia decades ago?

"why did the US move the industry"

An excellent question, in light of Sematech, which formed in the mid-80s to deal with the Japanese semi-conductor industry becoming a major source of electronics for the U.S. military.

Are there any examples of a government using social media to blackmail? And wouldn’t your argument mean China is justified in blocking Facebook?

It’s also quite dangerous to have the government decide what counts as “disinformation.” Every government justifies censorship as fighting “disinformation.”

Regarding semiconductors, we produce plenty for our own use. The only major importer of semiconductors is China (that’s why semiconductor stocks are so sensitive to trade developments). When the national security people say semiconductors are important to national security, they don’t mean we might not be able to get semiconductors; they mean the government might not be able to cut off China from getting semiconductors. This is not a legitimate national security interest; this is a desire to control the entire world.

I, uh, did not expect him to say PKD was the equal to Proust on interiority. Very interesting, though.

The author suffers from Industrial Policy disease. That someone who believes that Governments have good insights into the future economy and can successfully steer the economy in the right direction. This idea has been proven false in so many ways, it is kind of cute that he still believes it. Also, as usual for a wannabe technocrat he disparaged satisfied consumer, to him the most important thing isn’t happy people it is a strong state. Not my kind of interesting thinking.

More humble thinkers please Tyler!

Asia is notorious for government not only steering their economies but outright picking winners and they seem to be competent at it. Tyler's State Capacity Libertarianism endorsed China and Japan both of whome are adept at playing the heavy hand rather than the invisible one.

Japan stopped converging far short of US or even Western European living standards and China will most likely do so as well. The East Asian model has not produced a single country that successfully converged to US levels of income per capita, much less income per hour worked (people work harder in Asia)—the only “countries” in Asia that have successfully done so are Singapore, Hong Kong, and Macau, which followed more free market policies and are city-states to boot.

The China/Japan model seems to be a decent one for developing countries trying to accelerate convergence and don’t have adequate private financial markets and entrepreneurs, but it is simply not applicable to rich countries, where the economic foundation and stability for the free market is in place.

I agree with all three of these statements, although I would modify the first one a bit: yes it's highly questionable for a nation's government to involve itself too heavily in industrial policy (at least in rich nations, as Zaua points out).

But if you're Facebook or Google or AliBaba or TenCent then it is not only relevant but strategically vital for your company to make decisions and tradeoffs between how much to invest in social media algorithms and how much to invest in semiconductor manufacturing (or whatever hardware they work on, server farms or whatever).

So although Wang was thinking in terms of entire country's industrial policies, these sorts of discussions about where to invest are very much applicable and important to companies that are large enough to face those choices. If you're a big vertically integrated conglomerate you care about all of those items: it's not industrial policy, it's your company's business decisions.

Japan has similar GDP per capita numbers as UK and France while South Korea peers with Italy and Spain. Taiwan with Portugal and Czech Republic. It is the US that is an outlier for having both a natural resource advantage that the Asian Tigers lack and for being the incumbent superpower from the last century. If you subtract from the US the coastal, entrepreneurial parts of America the numbers come back to earth and reveal the stagnancy of the rest of the country.

I had forgotten about the modularism movement and the role of the Ordnance Department in American industrial history that Dan raises. The knock-on effects for all manufacturing of contractors gearing up for interchangeability seems to be left out of the article in the previous post. In particular, it likely externalized network effects that otherwise would have only benefited a dominant manufacturer's supply chain (and thus likely a specific geographic location as well) in any given industry. Together with an unusually well educated population, that may be enough to explain why American industrialization took a different path in the late nineteenth century.

The Bloomberg article referenced by Dan is pretty damning of Hong Kong:

"Hong Kong’s business landscape remains largely unchanged – the preserve of a small body of property developers and conglomerates"

"one of the most striking things of the city’s history for nearly three decades has been its failure to produce a single major new business."

"Hong Kong’s late-colonial features: a low-tax, capitalist economy, externally very open but domestically protectionist"

In other words, an oligopoly/monopoly led private sector with low taxes, free trade, and a big banking/real estate sector could still fail compared to Shenzhen which actually has many industries outside of finance. Let this be a warning call to NYC and London or any other so-called financial capitals that the real production of goods and services is more important to a healthy economy than financial engineering and other forms of rent-seeking.

Well it’s a magazine article. Overall HK seems to be a pretty good place to live based on GDP, longevity and access to amenities statistics. The main thing lacking seems to be political freedom, which is hardly the fault of capitalism. Of course house prices are high as well, but nice places always have higher house prices than not so nice places, I would suggest the answer is to make more places nice rather than try to change the ones that are successful.

Wang: "As a rule of thumb, it’s harder to name global Chinese brands than Japanese and Korean ones, even when they were close to China’s current level of per capita GDP. Shouldn’t we expect more from the world’s second-largest market?'

China's GDP per capita is where Japan's was around 1970. That year, Toyota exported 200,000 autos to the U.S. and 400,000 autos in 1971 so just 2% of autos sold in the U.S. in 1970. Honda first sold the Civic in 1972 and exported it in 1973. Sony's radios were well known by 1970 but an expensive product.

South Korea's GDP per capita was at China's current level in 1995. Those in tech would have known Samsung then, but I don't think well known in the U.S. Hyundai wasn't that well known in 1995 with 0.7% share and didn't have a better reputation until the 2000s.

China doesn't seem to be unusual compared to Japan and Korea.

1995 was the year that the CEO of Samsung took a giant pile (150,000) of company products and burned them in a big bonfire in front of the employees to emphasize the need for higher quality and less defects.

I meant: "Sony's radios were well known by 1970 but an *inexpensive* product."

Dan needs to look more at the current competitive situation regarding microprocessors: Intel is in the worst position they've been in 20 years, when the Athlon was competitive. Their foundries don't have a tech advantage, their products don't have a price advantage, and their products for machine learning aren't competitive.

They aren't dead, but the competition has improved so much, they could face a similar fate as HP's printer division

Meh, Intel doesn't really seem to be in much trouble. They still produce cutting edge chips. Furthermore, they have regained the position as the largest semiconductor supplier on the planet.

"Nov 18, 2019 - Intel to Reclaim No. 1 Semiconductor Supplier Ranking in 2019. With a forecasted 34 percent drop in the memory market this year, Intel is once again expected to rank as the largest semiconductor supplier and have sales that are 26 percent larger than Samsung in 2019, according to market research firm IC Insights."

Bob is talking about the technical situation of Intel so you can't refute that with market press releases. The fastest desktop and server chips are trending toward AMD who makes their chips in Taiwan and have already beaten Intel in some benchmarks for a while now. This is due to TSMC's new process advantage and because Intel suffered major hardware security flaws whose fix requires removing certain optimizations which in turn slows performance.

Intel may flex their market muscle on the wider ecosystem in the short term but that only gets you so far. In the long run, higher prices and lower performance will kill you in an industry that very much cares about both. They fired their CEO earlier last year (it's now 2020!), we'll see what their next move is.

Comments for this post are closed