Just the opposite. Need to let people raise prices to encourage slack in the system. Let’s prices settle where chip makers can produce new chips.
There are market failures where the government needs to step in, but this isn’t one. Even with climate change (where they should step in) the government can’t get to the point of saying it’s ok for gas prices to be high.
We don’t want the government to pick winners and losers when it doesn’t need to.
It's a puzzle to me why companies like nvidia didn't just raise prices and instead left both customers and profit to the dirty world of scalpers.
I guess it's "brand damage" but I feel there would be something more fair and honest if in times of tight supply they ran their own ebay-like store and auctioned them off. It wouldn't feel like a price hike and prices could automatically settle as supply/demand reaches parity.
Some, like this MSI subsidary, were actually selling on ebay at scalper prices.
> MSI has admitted that one of its subsidiaries has been selling RTX 3080 graphics cards on eBay at almost double the MSRP.
> The controversy first appeared on Reddit, where users accused MSI of scalping its own RTX 3080 graphics cards on eBay under the name Starlit Partner. Since, it’s been confirmed in a Justia Trademarks listing that Starlit Partner operates under MSI Computer Corp and was first set up in 2016.
They probably did, the only cards they produce under the Nvidia brand are the Founders Edition cards and those are sold for MSRP. Everything else is made by their partners and those cards increased in price.
Yes, soon after posting I realised I was referring to Asus/MSI/Palit etc. rather than Nvidia. It means either two levels of auction or the Nvidia chip sale is a % of the end unit sale.
It's a Taiwanese company around since the 90s. They may be more Europe focused with a German office. I've never heard a US reviewer mention them but in the UK, they offer cards cheaper than the other brands. Beyond that, I don't know much about them. I've bought a couple of their cards and never had an issue so they get a thumbs up from me for N=2.
To over simplify, let's say that you have 100 people in the market: 90 people who can afford $100 per card and 10 people who can afford $200 per card. In typical times, you charge $100 per card and sell 100 cards, earning $10,000 in revenue.
Now you have a chip shortage, and you can only produce 50 cards. If you charge $200/card, you only sell 10 cards, earning $2,000 in revenue. If you charge $100/card, you'll sell all 50 cards, and earn $5,000 in revenue. So it can still make sense to keep the price lower if it makes you more revenue overall.
What you've described is nothing like a real market. Where are the people in your model who are happy to pay $110? $120? If there are 10 who'd buy it at $200, and 100 who'd buy it at $100, surely there'd be 50 who'd buy it at $130 or so.
That ignores the social aspect entirely, too. How many who were originally willing to pay $100 will later pay $200 when they see others pay that amount for the item and it becomes scarce?
1. Your comment doesn't even provide a prime facie argument against some forms of government intervention. E.g., we could let the market determine prices but mandate that fab equipment suppliers go to the front of the line.
2. Price is probably a red herring anyways. I'm willing to bet that fab equipment suppliers are losing out not on price negotiations, but on volume negotiations. I.e., they might even be willing to pay more -- even much more -- than other users, but can't buy in massive quantities so don't go to the front of the line.
3. Is there any (legal) mechanism at the moment that prevents chip makers from increasing prices?
4. Fab equipment producers are small consumers of chips but have such a disproportionately high impact on the rate of future supply. In the midst of a global shortage, we could straight up socialize 0.00...01% of chips produced every year and hand them out for free to fab equipment manufacturers without even effecting the short-term price dynamics. I'm not actually advocating this, but the assertion that earmarking a small number of chips for a particular high-value use fundamentally skew the market in the short-term is probably false.
5. Even if markets can eventually work in this case -- and for the record I'm convinced that this is a perfect example of contract negotiators being extremely myopic -- market dynamics have non-O(1) time complexity and the chip shortage is wrecking havoc on the real economy.
My comment wasn't suggesting price controls or socializing chip fabrication. It was suggesting that we very temporarily give special treatment to a very small consumer of chips that has an outsized impact on production rate, in the midst of a global chip shortage.
>It was suggesting that we very temporarily give special treatment to a very small consumer of chips that has an outsized impact on production rate, in the midst of a global chip shortage.
Why would we do that if the fabs themselves don't think it is worth paying their equipment manufacturers enough to afford their own chips?
Giving "special treatment" is a price control. It is forcing a transaction that otherwise wouldn't settle at that price.
> Why would we do that if the fabs themselves don't think it is worth paying their equipment manufacturers enough to afford their own chips?
Because there is a global chip shortage that is making life substantially worse for the vast majority of Americans. And because markets are tools used by man, not the other way around.
Specifically, prices on new and used cars that most Americans depend on -- for better or worse -- to do basically everything in their lives (including getting to work, getting to school, getting food, etc.)
> Need to let people raise prices to encourage slack in the system
You're assuming that increased profit margins will automatically increase supply chain buffers. But rather, the same incentive to hire too many financialists that "save the company money" will exist, and the extra profit margin will just go to increased dividends.
And increasing the supply chain buffers won't help much right now either, it has to be done during good times. In fact I'd say most of the shortage is from companies deciding to increase their buffers, in the same way as what happened to toilet paper. "Hoarders" and "speculators" are easy illustrations to point to, but the real demand comes from regular consumers silently buying twice as much as they usually do.
> You're assuming that increased profit margins will automatically increase supply chain buffers.
I read GP's comment as increase prices to decrease hoarding, which in theory could provide the slack in the system. Problem might be that certain products may not be viable if prices get too high. Only those with sufficient margins prior.
I don't see that raising prices would make consuming companies want to stock up less. A company that requires an input to manufacturing is going to have a pretty steep price-demand curve for every individual component. Even if prices for one of their scarce inputs doubles, they're still going to buy as much as they can rather than risk shutting down their entire production if they run out. Since their competitors are feeling the same pressure, they'll just raise their own prices to compensate.
It's like foodstuffs during the early pandemic - when you finally found something that they had been out of, you didn't particularly care about the price, and you generally bought extra so you wouldn't run out if it went missing again.
> they'll just raise their own prices to compensate.
That's how it works in econ 101, but not necessarily in practice. Prices on many goods are less flexible than commodities like oil and lumber, for many reasons. Manufacturers may be locked into fixed-price contracts or distribution agreements, for example. Or a scarce component might be shared across "budget" and "premium" product lines, but the budget line is too price sensitive to change so the premium product goes up 10x instead. Or the company just borrows money and eats the loss...
Sure, but the alternative is for a company to shut down its production line rather than pay extra for parts it needs. None of your examples would seem to include that happening.
I agree they'd probably pay extra for the parts, but that might not just get passed on in their system price. It can manifest in all sorts of other ways (debt, discontinuing a different product line, cutting back on research, etc).
In today's environment debt is cheap, so companies that might otherwise shut off a production line can afford to borrow and bid up the price of parts.
"Just the opposite. Need to let people raise prices to encourage slack in the system."
"We don’t want the government to pick winners and losers when it doesn’t need to. "
Markets are not even remotely close to as efficient as you're implying.
In a clinch, people are making all sorts of crazy guesses at what the future will bring, making everything very inefficient. Remember that efficient markets depend on rational acting based on good information. We often don't have very good information at the unit level, and, we often act irrationally.
Some company flush with cash, decides to buy things at crazy high prices thereby denying the 'critical sources' (those that support production) access.
Right now there is a lot of parts hoarding - speculators buying up parts to sell them at higher prices. They're adding no net value to the system and causing all sorts of other problems.
The clearing of those prices may happen over time, but not without terrible damage being done.
Supply chains are not like stock markets with clear prices and instant transactions.
You may not need the government stepping in, but you definitely want non-market actions. For example, chip makers may want to work with their supply chains to ensure a kind of absolutism or preferential customer tranches.
FYI this already happens, all the time. Price is not King for parts, like it is on the stock market. Vendors of 'everything' are aware of the long term growth of their business, and will generally want to work with consistent buyers.
So in this FUBAR panic, supply chains have to think not about one thing, about many.
I believe that root causes was already a fait-accompli at the start of the pandemic when a bunch of parts of the supply chain shut down - we're still paying the price of trying to get things going.
What makes you think the most critical uses have the most money to pay for chips? I'd imagine there's a lot of crappy (or in any case not critical) consumer products that have much higher margins.
There is more money in consumer markets because there are more customers, not because each customer can pay more. It's usually businesses that spend big on individual purchases.
> Need to let people raise prices to encourage slack in the system.
The scalpers can absorb the rising prices until the desperate companies no longer can afford the increase. That will cause the market crash, which is not good for anyone.
I think government should regulate that space so that businesses engaged in scalping could no longer purchase nor sell the chips.
>The scalpers can absorb the rising prices until the desperate companies no longer can afford the increase
The whole point is for manufacturers to raise the price until desperate companies/consumers can no longer afford it and don't buy them. Im not sure where scalpers come in. If prices are set high enough, scalpers can't make a profit.
It does have negative external effects, as you observe, in the sense of a viscious circle. But even a viscious circle, mathematically, has positive feedback, resulting initially in a positive exponent growing against time, not a negative exponent damping out in time.
Doesn’t positive feedback require a positive loop gain? Right now, not enough chips are available to make new chips. So the loop gain is less than zero, damping the output recursively.
I have always understod "positive feedback" as "feedback that prompts the existing change to continue in the same direction, with equal or larger speed".
So positive feedback on a falling signal would tend to make it drop more. And positive feedback on a rising signal would cause it to rise more.
So, basically, for any signal S at time t, we would expect something like S(t+1) = S(t) + S'(t) * k, for some k > 0. And blatantly abusing derivatives for "should probably be a delta between S(t) and S(t-1)".
But, then, I am not a control theory specialist, I don't even play one on TV.
The problem is, if the feedback signal is too slow in getting back to the thing which measure error (i.e. too much phase lag), then negative feedback can turn into positive feedback and this leads to an instability.
One of the most complex pieces of the semiconductor fab is the building itself. Even with plans and permits in hand, it takes years to make one that can output at reason throughput and yield.
This report is from 1999 and it hasn't gotten easier.
"Typically the product life of a semiconductor chip (nine to 12 months) is less than the time required to construct the facility and install the equipment for manufacturing (24 to 36 months). As such, the construction/commissioning process is a rapid, constantly overlapping and complex set of events. In addition, construction of semiconductor facilities is very complex and costly (about USD 1.2 to 1.5 billion) due to the extraordinarily sophisticated processes and equipment required to manufacture semiconductor chips."
"Typically the product life of a semiconductor chip (nine to 12 months) is less than the time required to construct the facility and install the equipment for manufacturing (24 to 36 months).
That's an absurd underestimate of market lifetime. I'd bet that fully 80% of the chips available in 1999 when that report was written are still in production today (or would be, if not for the crunch.)
Yes, and same situation in every mining-related commodity market. Multiple time delays of order several years. Large up-front investments. Large uncertainties in payoff. Look at the multi-year price behaviors in those markets and see too if there is much stability.
I'm also confused by the confusion here... probably naive pattern matching? Reminds me of TAing undergrad courses where you could get more than half the class to confuse "positive feedback" and "negative feedback" on a midterm by just giving examples where stability = bad :)
Grandparent gives the choices of ‘positive exponent growing against time’ or ‘negative exponent damping out against time’. Here we have a positive exponent damping (reducing) the output over time, because it’s value is less than one, resulting in a negative loop gain. The Wikipedia article linked up thread defines positive feedback as having positive loop gain, and negative feedback negative loop gain.
IIRC, generally a positive loop gain greater than one will lead to diverging behavior aka instability, whereas even a positive-sign to feedback, if loop gain is "less than one" will not. I might be brain-farting here, but I cannot be precise anyway, which IIRC gets into plotting poles in a complex plane. (Cue joke my applied math professor would tell, about why all the Polish people were asked to sit in the right-hand aisle of an aircraft.)
Negative feedback means the feedback coefficient is negative, you are trying to redefine the usual meaning.
It is like saying that I tested positive for COVID-19 WRT my health because I don't have the virus, it is not what a positive test means and it will confuse anyone who knows the correct terminology.
No! It's negative WRT chip production rate! I have a phd in controls and meant what I said; I assume roboticsresearcher also knows some freshman-level control theory ;-).
Stability ==== good in undergrad engineering, but not here. We DON'T want production rate to be stable when we have a global supply shortage! Here, a negative feedback loop is stabilizing the system in an undesirable equilibrium.
I.e., the function that's being controlled in "supply of chips", the stable state is "saturated supply", and the negative feedback loop that maintains that equilibrium is "starving chip fab suppliers".
(meta: people down-voting comments on control theory terminology by two different experts in this field at least makes me feel a bit better about the signal:noise ratio on the vote counts on my other comments in this thread ;))
Only partially facetious here: you have mistakenly decided that it it clamps at zero. If we consider the hoarding aspects, the feedback continues into net negative chip availability. Before long, there will be roving gangs of looters taking back the chips you thought you already had! ;-)
No, there is a shortage, leading to a shortage, leading to a shortage, ...
If it was negative feedback, the "error" (shortage in production) would lead to an error cancelling signal, and therefore an increase in production. Positive feedback has error leading to larger error, shortage leading to more shortage.
What experiences with which government would lead you to believe that bureaucrats -- most of whom have never run a company nor made a product -- would be capable of "managing the market"?
I'm not sure experience running a company is necessarily a good thing here, or at the very least, not pertinent. Wall St.'s and the American people's interests are not necessarily in line. The US' previous president ran some companies and did not really do a good job in this regard, either, wrt his trade wars.
You're simply wrong. The US's economy was doing _great_ under the previous administration, and even top Democrats agreed that it was right to put strong pressure on China re: trade.
Great for whom? Wall St. or the common American? Like I said, their interests don't necessarily align. Sure, maybe stocks were up (usually, not when Trump was threatening to shut down the government if he didn't get his wall, though), but that didn't trickle down to everyone and is far from a total picture of the economy. Also, your article doesn't support your claim. Schumer was praising Trump for being tough on China for the sake of being tough on China, not for managing supply chains well.
Seriously? You're using Donald Trump, who specialized in brand licensing and being a television character, as an equivalency to all the manufacturing engineers and supply chain specialists working to resolve the problems created by a global pandemic that's killed millions of people?
The amount of disrespect to highly skilled professionals in this thread working like crazy to respond to a massive exogenous shock, and then following it up with the idea that "well, the government should fix it" with no specific idea of how exactly, the government would fix it, is mind-bending.
This is a pretty axiomatic view that governments can never be as responsive or as efficient as markets. The more commonly accepted economic wisdom is that they are usually less efficient and responsive than free market forces operating under ideal conditions. There is a lot of room for market failures, inefficiencies and temporal dynamics to change the balance. The are plenty of examples of government regulatory bodies that have a nice anchoring effect on the relevant markets. The Federal Reserve Bank, for example, responds quickly to changing market conditions, looks at the data and intervenes at a speed that keeps pace with the rapidly fluctuating market it regulates.
The Federal Reserve hardly responds to market conditions. Their free and open printing of money is arguably the greatest risk to our economy. Milton Friedman has great resources on this, including a video series from the 1970s (based on clothing alone) called Free to Choose. Its on Youtube, amongst other places.
If republicans think that unemployment benefits compete with private businesses on labor, then I get to think that 0% interest money competes with labor for capital.
The fundamental problem is that the 0% lower bound combined with a deposit guarantee represents not only a minimum wage for capital. It also presents a job guarantee. A minimum wage doesn't guarantee you a job.
So yes, the Federal Reserve is not responding to market conditions at all. It's artificially holding up interest rates at zero or above. This is causing massive distortions in the economy that can only be fixed by a swiss-army knife of policies. Among one of the needed responses is "free and open printing of money". The world economy is already flirting with disinflation (a reduction in inflation). If you don't have negative interest rates you will need a whole load of "money printing" to keep the system standing in place.
The assumption that a scarce money system (i.e. guaranteed non negative interest) has a fixed velocity of money is absurd. Put interest at -5% and just watch everyone withdraw cash from their bank account. The velocity of cash would be basically be zero and the velocity of money on bank accounts would be extremely high. As the government is doing deficit spending all the money just piles up somewhere and ends up doing nothing. QE is even worse because you cannot spend centralbank reserves to buy groceries.
Ok, let's do the negative interest thing. It sounds like a big hassle right? Just think about the benefits: The first step after negative interest ratess would be to adjust the inflation target to 0% meaning perfect price stability. Actually, you wouldn't target inflation at all because the negative interest rate completely replaces the need for inflation. You would target the CPI itself meaning your goal as the government would be to maintain a CPI of 100 for all eternity. Any deviation would become inexcusable. Meanwhile today inflation is a hack to make a broken money system work.
Natural boom-and-bust cycles are a huge risk to economies as well, despite the fact that they're natural and don't have anyone for which we can point a finger at.
I don't know what you mean by "natural boom and bust cycles" but credit cycles and other financial cycles are just a property of the money system. Alternative money systems do not have this property.
The cycles only make sense because people like and want them. I.e. they love the scarcity of money. For example, in a depression the return on money is greater than the return on labor, people logically flock to money rather than labor even though real wealth is eroding as people stop working.
> The Federal Reserve Bank, for example, responds quickly to changing market conditions
The Fed was established by Congress and the Chair is appointed by the President, however the Fed is still a private institution. That independence makes it a very different organization than what most people mean by government.
It's unrealistic to expect people to hedge online comments about complex topics (like the one in question here) to the extent required to preclude "yes but" and "well akshually" type comments that complain about the lack of nuance. Yet despite these expectations being unrealistic everyone expects comments they disagree with to meet them.
We're in this mess due to our never ending quest for more "efficiency"
The government needs to recognize the fact that semiconductors are essential to national security and ensure we have the capability to produce our own.
Then what's the solution? Seems like the government is the only entity that can intervene if market forces are counter to supply chain resilience. This is partially the justification for agricultural subsidies in the US, so clearly there is precedent.
Toyota pioneered the "Toyota Way", which is now known as Lean or JIT manufacturing. JIT is famously susceptible to disruption from natural disasters. Over the short term, between disruptions, JIT tends to be more profitable than the alternatives. Over the long term, the market rewards companies that can handle disruptions. Basically, to answer your question, the market is punishing JIT MFG and rewarding resilience. Companies are watching it happen and learning from it. One indication of this is the current increase in inflation. Companies are switching from 1 month of inventory to 6-12, which is making suppliers scramble and driving up prices.
I am by no means against government intervention. Companies have short memories, and market forces will force eventually pressure a return to JIT. But now is the exact wrong time to intervene.
Time and capital investment. It's like this generation of people have never heard of production and supply disruptions, and were oblivious to such things being possible. Frankly, this doesn't matter very much, it's not a critical situation.
The auto market malfunctioning short-term due to a pandemic doesn't present a strong argument for government intervention. Tesla can't make batteries fast enough, there isn't enough supply, its restraining their auto production, the government must step in and fix the problem! It's nonsense. The government should not step in every time there is a short-term problem in a market.
Toyota won't sell as many vehicles. So what.
I know, I know, but what if people have to make due with a three year old vehicle. What if they have to suffer and endure those vehicles being made to last for five or six years. Ten years! The horror.
Toyota won't die. Time will pass, during which necessary investments and adjustments will be made. Supply will be increased. The problem will be fixed. It's as simple as some time and capital investment. The companies that maneuver the best will come out ahead, gaining an advantage on their competitors. And the world keeps on spinning.
Toyota has generated something like $90-$100 billion in operating income the past five years. They have the resources - and then some - to fix the problem. If they choose not to or can't that's their own incompetence, their competitors will eat their lunch. Never feel bad for a corporation earning $20 billion a year. If they can't get their production corrected, someone else will figure it out and reap the benefits.
It does not matter as much as is being portrayed. This is not an important problem and does not warrant the government burning its time and resources to step in and fix (assuming they can help at all). Governments have a lot of other far more important things to be focused on.
The U.S. government has decided that it's in our national security interest to remain a net exporter of crops.
If World War III broke out and all the borders shut down, America would still be able to feed herself. The U.K. wouldn't. There would be mass starvation in much of the first world, and people would say "the government should've done something."
All the diabetes is a pretty rough unintended consequence, I'll give you that, but shifting some chip fabs to our shores as a matter of national security doesn't sound like too bad of an idea.
While I agree there are definitively some downsides to ag subsidies, I think the real question is if the interventionist downsides are worse than the non-subsidized downsides. As bad as they are, I'm not sure that incentivizing unhealthy food is actual worse than famine.
Cratering agricultural prices so margins are so thin small farmers are driven out of business in favor of huge conglomerates, mind-boggling levels of food wastage on the order of billions of pounds sitting in warehouses until they rot.
I agree, those are all blowback of subsidies. But they are also side effects of food abundance. The downsides of food shortage seems much worse.
Maybe there’s an argument that we’ve moved passed the era of food scarcity when those policies were enacted and they should be modified. But I think a blanket claim that food subsidies are an inherent bad policy misses their point.
That blanket claim was never made. The made claim was that the ineptitude and mishandling of agricultural subsidies is reason to reconsider calling for government involvement in the chip shortage.
This is a bit of a confusing take if you're using ag subsidies as evidence to keep the govt out of chip manufacturing but now denying that the subsidies aren't bad.
>The nightmarish results of agricultural subsidies in the US is an excellent reason to not involve the government
This sure sounds like you think it's a claim of subsidies being bad policy.
>The made claim was that the ineptitude and mishandling of agricultural subsidies is reason to reconsider
How do you combine the view that "ag subsidies aren't bad" with "the government shouldn't be in the business of managing subsidies" when the definition of subsidy involves the government? At first take, this comes across as back-peddling to avoid dogmatic cognitive dissonance.
But I'll be generous and assume you did not mean that ag subsidies are bad in and of themselves, but the way they are handled is poor. So what do think is a more proper way to handle them? Should the focus be on different products? If so, which ones?
The point has already been made that ag subsidies are operating as intended and the downsides you refer to are downsides of abundance. I have a feeling that most people who have actually lived with food scarcity would find them preferable to the actual "nightmarish results" of too little food.
Subsidies in theory can work for some problems. My argument is that the US government is not capable of executing them in a competent manner. The collapse of family-run farms is a "downside of abundance" in the same way drowning is a downside of abundance of water.
Thanks for clarifying, although I disagree that a collapse of family run farms (while bad) is worse than scarcity of food, both at the personal and national security level. I personally wouldn’t want the government to necessarily optimize for “number of family farms” at the expense of other concerns
That's a false dichotomy. The choice isn't between food scarcity and food waste but between a normal industry and decades of enormous waste. Experts were complaining about the subsidies damaging the dairy industry all the way back in 1983 yet those same subsidies are still here forty years later. The US government is happy to smash a problem with a hammer made of other peoples' money but they have zero ability to foresee the consequences of their meddling or to stop themselves from continuing to smash long after the need is gone.
I’m not claiming a dichotomy in that family farms and food abundance cannot coexist; I’m saying you are using a metric (family farms) that is not the intended goal (national food security). So again, I’m asking what is the better alternative for implementing subsidies?
I don’t know if we’re just talking past each other, but it’s hard to make sense of your stance. If you think subsidies are a viable solution to some problems but the US govt can’t manage subsidies, are you implying the US should not use subsidies, even on the problems they would solve? If so, can you elaborate on govts that have used subsidies to solve similar problems by better implementation? When govt is, by definition, who wields subsidies these are difficult points to reconcile.
Pointing out less than perfect implementation isn’t really helpful unless you can figure out a way to improve it. Just saying “the govt is inept” isn’t helpful when they are literally the only organization who provides subsidies.
I don’t disagree with the problems you point out but they come across as flippant “see!? See how bad the govt is!?” Dogmatic axioms might make someone feel good without actually addressing the problem.
I’ll give an example. I think subsidies need to have clear metrics to measure effectiveness and sunset clauses. This would help prevent things like alpaca subsidies meant to assist in the Korean War somehow staying in place until the mid 1990s.
The point of the subsidies was national security, not “protecting the family farmer” or “minimizing food waste.” To that end, they worked.
The comment was a response to a suggestion that the solution to the chip shortage was government intervention. I don't consider giving a evidence to the contrary of an argument as pontificating.
Ok but point to the part of that comment which was evidence?
Claiming that the results of ag subsidies have been "nightmarish" with no further elaboration or citation does nothing to advance the conversation, it's simply a strongly worded opinion.
That's one way to look at it. Another is that those are side effects of having stable and affordable food prices. The solution being talked about in this thread of simply raising prices would literally starve people to death if applied to that case.
In direct opposition, Auto makers approach of minimizing inventory and producing "just-in-time" caused them to be vulnerable to supply chain or big market shifts
The article notes that Toyota avoided "just-in-time" supply for chips, and has benefited for a while from this. Their stockpile just ran out.
> New cars often include dozens of microchips but Toyota benefited from having built a larger stockpile of chips - also called semiconductors - as part of a revamp to its business continuity plan, developed in the wake of the Fukushima earthquake and tsunami a decade ago.
The trend IS NOT towards diversification. In the last 5 years since Trump's election, US multinationals were increasing their presence in China, not decreasing.
Google for example said to open "a small representative office" in Shenzhen 2 years ago, now it's a full giant RnD centre in the Ping An Tower where they shipped all of Pixel's development.
Apple had RnD offices in China for more than a decade, but they barely acknowledged their existence. Their people in the Kerry Plaza were prohibited by their contract to even show their employment for Apple in their LinkedIn profiles. Their Shenzhen RnD centre is where AirPods were developed, along with many other iPad, and iPhone sub-assemblies. Apple's VR goggles project had its start in Shenzhen as well.
Amazon had no presence whatsoever in China besides a failed Chinese Amazon.com launch. They left China, and then returned to move the whole of their Kindle, and Echo device development to Shenzhen. Now they are working on something rather cryptic there. Some suggest VR goggles of their own design.
Facebook... absolutely bizarrely opened their RnD centre in Shenzhen amid the COVID, just a floor below Google I heard.
Dell, Microsoft, Nvidia, Qualcomm, Intel — all conventional hardware makers were here since nineties, but I think they really doubled down on China recently as well too, to one up the dotcom upstarts in hardware.
Having your essential RnD office shut down, if something happens to/in China, would not be any much less ugly, and disruptive than having your access to microchips shut down.
In other words, the Silicon Valley is still going all in on China, despite 4 years of Trump, public scorn, trade war, rising costs etc.
In other words, they really gave up on any vision where they don't critically depend on China, and can run with critical assets in US only.
That sounds nice, but a singular entity within a government, entrusted with the necessary power to regulate the relevant parts of the market without too much coordination overhead with other government entities, is actually much more efficient at resolving market inefficiencies than the free market. Case in point: production of vaccines.
That's also obvious: the inefficiency in governments originates largely from coordination overhead between many competing entities with overlapping responsibilities. Self-regulating systems like markets do not eliminate that overhead, they just use other means of coordination that trade some of the complexity overhead for a time overhead - instead of having to coordinate a complex set of rules, you now have to give the system enough time to "find" its stable state. But when time is of the essence, an intelligent, singular entity without the need for coordination with anyone besides the entities to be regulated can always outcompete the self-regulating system when it comes to short-term stabilization (though not necessarily with regard to long-term stabilization, but that's not the issue here).
The US government regulated critical parts of the supply chain of raw materials and preproducts in order to ensure that US manufacturers have no sourcing problems. It then compelled the manufacturers into exclusively servicing the US purchase contracts first before fulfilling competing contracts from other global buyers with any of the finished product produced on US territory.
I'd call that quite a lot of "market management". But as everyone could see it resulted in the fastest vaccination ramp-up worldwide (excluding Israel, which was a bit faster, but is also much smaller than the US and which had its own way to get "preferred" access to vaccine produced in the EU).
This assumes an efficient global market, without interference from other state actors.
If other countries are implementing protectionist policies, like keeping semiconductors for themselves or supplying other nations first to curry favors or improved relationships, for example, it might be in another nations interest to increase fab capacity with its borders to avoid being vulnerable to those political and diplomatic factors.
DARPA funded the basic research. Then the actual Internet was largely built by for-profit private sector companies. That hybrid model seems to usually be the most effective for major new innovations.
DARPA funded most of the grant money used by research entities to create the various parts of the IP stack. That's not "basic research". Simply put, were it not for DARPA, we wouldn't have the Internet of today. We'd probably have something like it, but something a lot more closed off and walled garden-esque.
They can't even manage a trivial task like maintaining roads and bridges properly.
You could have hardly picked a worse example than the government roads system (including our thousands of dilapidated bridges), which is in absolutely horrific condition and is a humiliating example for the government. It's the opposite of a good example.
That's all due to lack of funding, one might suggest? They're not lacking for funds. They spent our money on blowing up other countries and then (occasionally) attempting to rebuild them. Vietnam, Iraq, Syria, Libya, Afghanistan, Korea and 497 other cases of meddling and foreign adventurism. There rests $10 trillion in infrastructure money. No, they have had plenty of our money to spend, and they chose to squander it.
What ever would we do if we didn't have those hyper incompetent clowns to manage our roads.
GPS is trivial. If Russia can do it, various US private corporations could easily do it just the same.
US private corporations maintaining roads implies a society like the one depictive in the Robocop movies. It's fun to see you prosing that unironically.
While a lot of industry seem obvious and worth the investment today, when they were nascent that wasn't the case. Would GPS exist if private companies had to fund the rocket and satellite research just to tell you where you are on the map? Or would the aircraft industry exist?
Most of those types of high-risk, nebulous reward (at least on short-to-near-term timescales) industries are predicated on government investment. SpaceX, as great as they are, probably wouldn't exist if they didn't have NASA as a customer.
it certainly does have hindsight bias, on both sides. The Satellite TV and Radio industries spend the money to launch.
Back in the 90s (during the age of early GPS), Motorola (iirc) tried to launch a satellite phone company too.
NASA isn't that big of a customer. DirecTV was a huge consumer of rockets. As it turns out, SpaceX doesn't sell to them because they own their own rocket company.
>The Satellite TV and Radio industries spend the money to launch.
Correct, but this is an after-the-fact understanding. They spend money to launch now because the industries are no longer nascent and being launched on platforms designed around government investment. The key to my point is that the government spends money when the industries are young and risky to develop platforms. If those platforms work out that helps the private sector have a less risky path down the road. This is why telecoms weren't rushing to develop rockets in the 1960s.
>NASA isn't that big of a customer.
This is very much the same thing. Early on, NASA was really the only SpaceX customer and NASA helped keep them from going bankrupt [1]. In addition, NASA made early launches more palatable because the government is self-insured. Government contracts help usher along young, risky companies until they could have a less risky business model that the private sector feels more comfortable with. It's very similar to aerospace development over 100 years ago with the Wright brothers and Curtis vying for Army contracts. Without those contracts, they are as much hobbyists as entrepreneurs.
The free market can do a lot when it has the basic requirements to operate, but the point is that government spending is the only solution when we end up with a chicken/egg problem as the OP pointed out.
Why? Is the government better at inventory and/or supply chain management than the current players in the market? My experience with COVID tests says no.
No, but the government can shoulder a larger risk than what the private sector may tolerate.
There's all kinds of examples, but aerospace is a classic one. There would be no airline industry or commercial space industry if the government wasn't willing to bear a disproportionate amount of the risk when these industries were nascent. There just wasn't enough market demand to incentivize the private sector to do so on their own. So the govt sets up an incentive structure that brings the risk to a level where the private sector is willing to partake. The government is also generally more tolerant of longer-term scenarios than the private sector.
No one is saying the government should start producing chips, but they can offer financial incentives to encourage chip manufacturers to build fabs here instead of overseas.
> A negative feedback loop in semiconductor supply seems like a situation where a government should just step in manage the market a bit.
Unfortunately, no Western government can do it even if its life depends on it.
German trade officials for example went and completely prostrated themselves in front of Taiwanese govt, and TSMC, offering anything short of switching the recognition of China to Taiwan.
It bounced off without any effect.
It was only a blank cheque from USA that made them to even scratch, and that is still pending that cheque being honoured, and cashed out.
