Hacker News new | ask | show | jobs
by JavaBatman 1669 days ago
I am still amazed by the shortsightedness of how these critical and strategic goods are off-shore. No wonder why the supply chains are fragile and not robust. It should be illegal for strategic goods and components such as chips and semiconductors to be produced offshore. Yes, it will raise prices in the short and medium-term, but it will make the domestic economy and manufacturing sector more robust, resilient to supply shocks, and antifragile.

We need a true industrial policy and state capitalist system similar to what Japan, South Korea, Taiwan, and Singapore have.

12 comments

The blame for this is purely on the shoulders of business leaders coming out of MBA programs across the US.

It seems the only thing those places are teaching is how to join a company, strip it to the bone, hype up the short-term 5% growth, and get millions of dollars in raises.

Blaming individuals doesn’t help, it’s a coordination problem. If any one MBA refuses to outsource they’ll just be outcompetes by someone who does. This is the sort of problem that requires a government to solve
The blame is clearly on the shoulders of politicians and those who keep voting for them. Most of the political class are pushing globalization, and thus align tax incentives towards outsourcing. Imagine if the politicians in charge practiced what they preach and refused to buy or import goods that were produced by countries who don’t meet US labor law or emissions standards. The problem would practically solve itself.
>blame is clearly on the shoulders of politicians and those who keep voting for them

Or a system that depends on "voters" to know any of the nuances and details of manufacturing and supply chain importance, or actually vote with long term broad national interest in mind, instead of short term, out of narrow self-interest, and largely be ignorant of all manner of political, economic and other issues.

It would also collapse the economy into a depression never seen before, so that may be why it doesn't happen.
I don’t like those MBAs anymore than you but blaming this group is disingenuous and an oversimplification. Business schools train people to profit, and if that comes at the cost of secure supply chains and national security, who cares, it’s not their problem. Someone else mentioned that this is a coordination problem, and I think that’s more accurate. You need governments to help fund these capital intensive ventures and incentives like tax-breaks and the fear of supply chain destroying pandemics to urge them on.
And people wonder why China structures their political system to subordinate the interests of multinational corporations to the longer term interests of the nation, its people and economy as a whole.
To be fair, they are often just following orders: "Maximize shareholder value" with a quarter/year schedule.
Which is why ceding so much power to these entities, rather than vesting that in the state and the people, is a mistake. Of course capital free to come and go and flow wherever it wants will discard a nation and its people's interests to pursue global profit margins at every turn. Just letting that happen with no recourse and restriction is a silly way to run a country.
MBA programs do not produce "business leaders", except by accident. They produce business _administrators_, which is a different thing entirely, and often the opposite that of a leader. People often confuse management with leadership, and while a manager can be a leader, the two are rarely the same.
How does making something illegal counteract economic forces? Hint: Look at Soviet Bloc economics. They imported high tech goods even after spending loads of money on trying to build them internally.

I would agree that onshoring more of the supply chain makes a lot of sense, but I think it must be an economically viable solution. I don't know exactly how to do that, but things like having a guaranteed market (like we do for food) might be part of the solution.

No single country is capable of producing the newest chips end to end. Lithograph machines, raw materials, clean room builders, educated workforce, and the scale to do each of these economically are simply not available in one location.

By the time you could produce such a focus of the industry, it would already have moved on and require different inputs that are not in that country.

I think that's what the person you're replying to is saying. The government should subsidize domestic production because it's a prestige issue for the nation. Sort of like having an aircraft carrier group but in the form of economic power.
It's not that simple. The components of a chip go through something like 40-50 countries, and include tens of thousands of suppliers. Involved in this are some of the most sophisticated technologies known to man.

To centralize all of this in one country is most likely impossible. It's not even a matter of throwing money at it. $100 billion won't get you there.

> I am still amazed by the shortsightedness of how these critical and strategic goods are off-shore.

You're not half as amazed as Xi is.

> Yes, it will raise prices in the short and medium-term,

Right now we have cutting-edge fabs in Taiwan and Korea. If we were to follow your recommendation, we would also have them in PRC, USA, Japan, Germany, the UK, India, France, Italy, Canada, and Russia, at least. In fact, that's more or less how things used to be. The difficulty is that this would mean increasing world fab investment by a factor of 6, which would probably make them all unprofitable until chip prices also increased by a factor of 6. Permanently.

However, there is still one country that uses this policy: North Korea.

Taiwan, South Korea, and Japan don't have complete semiconductor supply chains either. All the companies in those countries are dependent on suppliers all over the region and the world.
> I am still amazed by the shortsightedness of how these critical and strategic goods are off-shore.

At what point do goods go from being non-critical/strategic to being critical/strategic? At what point did car companies have to worry about chips being "critical"? At what point did smartphones become "critical" to people's lives?

Things were not critical at all, and then when few people were looking, they moved across the spectrum to an area part that is considered critical by some measure.

semiconductors have been critical since the day they were invented, because computers were already critical at that time. the first major manufacturers of semiconductors were defense contractors.
Well, we would actually have to rein in the oligarchs and puncture the asset bubbles to do that, and I think we have 0 stomach for either.

Our public transit woes also demonstrate how little US politics cares about details, and how much they stomp all over the independence of a civil service that would.

This announcment by Samsung followed news that the onshoring semiconductor bill was moving through Congress (just saw it in the news today or yesterday). I doubt it's a coincidence,

So, I think, at least with regard to chips, your dream is coming true.

When you start thinking about politics/geopolitics from an "interest group" perspective, instead of a national perspective, you will understand all these. "Nation" is just a concept that we put up together as a "framework". It means a lot for common people like us because we simply have nothing else to rely one, but not much for different interest groups (I'm talking about banking, military, lobbyists, industrial conglomerate, those big shots).
You don't make semiconductors out of thin air. They're made out of inputs that we have to buy from other countries. The United States cannot possibly produce its own sufficient supply of rare earths and silicon for instance, they must be extracted where they occur naturally.
The US did produce its own sufficient supply of rare earths and silicon for decades, they occur everywhere, mainstream semiconductors don't contain any rare earths, and I suggest you at least skim a Wikipedia article on a relevant topic before writing your next HN comment.
The US imports massive amounts of silicon (from countries like Russia no less). We do not have our own supply.

And no, rare earths are not contained in semiconductors specifically, but I assumed the parent comment wants to onshore more production than solely semiconductors, and rare earths are needed for lots of essential electronics.

Yes, the US has its own supply of silicon. Toledo, Ohio, was called the "glass center of the world" in the 01930s because of the combination of cheap energy (natural gas), low freight rates, and the high-quality silicon-dioxide sand mined at Silica, Ohio, twelve miles to the west (10.2307/141587). The US is the world's leading producer of silica, producing about 100 million tonnes per year and exporting 4 million (https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-sand-grave...), ten times the 0.4 million tonnes it imports. The world's highest purity silicon dioxide deposits are at the Spruce Pine mine in North Carolina, and these are commonly used to source silicon for semiconductors, because you don't have to spend as much money to purify it; also they are used for high-purity glassware for silicon processing (https://www.thequartzcorp.com/ https://www.wired.com/story/book-excerpt-science-of-ultra-pu... https://archive.md/BMFtO).

More broadly, it's difficult to find a rock where silicon isn't a significant component. It is difficult for me to imagine the level of ignorance that could lead someone to claim that an entire country lacked silicon resources. It could be remedied by reading the first paragraph of the introduction to the English Wikipedia article about silicon (https://en.wikipedia.org/wiki/Silicon).

It turns out to be true that rare earths are used for lots of essential electronics, though most components are devoid of them. Essential electronics are silicon semiconductors (silicon, aluminum, copper, boron, arsenic, phosphorus), metals for wires and traces (copper, tin, lead, silver, gold, zinc), FR4 (glass fiber and epoxy), optoelectronic semiconductors (indium, gallium, phosphorus, arsenic), high-speed semiconductors (indium, phosphorus, gallium, arsenic), inductor cores (steel, silicon, barium, manganese, nickel, zinc again, cobalt, strontium), quartz crystals, and capacitor dielectrics. Capacitor dielectrics include plastics, mica, electrolytic anodized coatings, tantalum or niobium pentoxide, and ceramics. Ceramics include the high-capacitance ferroelectrics (lead, zirconium, titanium, sometimes barium) and the high-stability NP0/C0G paraelectrics.

And this is where you finally got something right! It turns the NP0/C0G dielectrics do often contain rare earths: oxides of neodymium and samarium. There are non-rare-earth alternatives made from silica, manganese, titanium, barium, and zirconium (https://patents.google.com/patent/US5599757A/en) or titanium and magnesium (https://exxelia.com/uploads/PDF/ceramic-capacitor-non-magnet...), but the rare-earth compositions are widely adopted. Perhaps they have slightly higher permittivity (permitting smaller capacitors) or lower costs. I don't know.

Regardless, essential electronics can be made without rare earths with only minor compromises, and of course rare earths are everywhere; they could easily be mined in the US again.

They don't use silica sand to make semiconductors, they use quartzite. Which isn't available near near Toledo. There's plenty of other places in the US, but first you have to actually know the right mineral to look for.
You say, "They don't use silica sand to make semiconductors, they use quartzite." Even if this were true, quartzite is just silica sand that has been sintered together (naturally, underground). Moreover, Vince Beiser's Wired article (linked above) claims that in fact both high-purity silica sand and lump quartz (probably, as you say, quartzite) are used as sources of silicon.
My understanding is that the US definitely _could_ produce everything needed, it's just not financially viable or desirable to do so. With tax changes, automation and the rise in cost of labor in non-western countries, it might just become viable.
> We need a true industrial policy and state capitalist system similar to what Japan, South Korea, Taiwan, and Singapore have.

You forgot China. If you institute strong controls over industry in a country the size (and culture) of America, you won't get Japan, you'll get China, and the corresponding anti-features of their system (such as authoritarianism in general, and companies acting as agents of the state in particular).

Very few people want that, and certainly not me.

There are far better ways to build robust manufacturing systems than complete government control.

There are some uncomfortable questions raised here. Why is it desirable to build "robust manufacturing systems?" Why was it desirable to overengineer appliances to last decades in the 50s, other than "the designers felt it was the proper way to do it" and there was no market pressure to steer them otherwise. How many resources were wasted on such appliances over those decades?

If the supply chains have very little slack and products are designed with minimal material inputs, sure there may be some availability hiccup here and there, but how bad is that really and by what metric. At the same time many more people get to have access to modern goods due to lower costs and less unneeded material, less is wasted due to inventory stockpiling, and we are overall more efficient.

Samsung is considered the epitome of state capitalism, so there may be some confusion about that last sentence.