Nine governors press U.S. lawmakers to pass semiconductor funding bill
reuters.com
reuters.com
Federal pork will not make our country better at semi autolines.
For national security, we need a state of affairs wherein it is profitable to build and run a plant in the U.S. that responds primarily to the dynamic needs of private actors. All of the issues there are on the cost side of the ledger.
It is a company that is responsible for the final integration of a system, but often fans out subcomponents to a large assortment of other companies.
The way to do that is massive tariffs, not pork. They will be met with retaliatory tariffs.
Seemed to work when Taiwan basically funded TSMC. If they can do it, why can't we do it?
I mean, the USA is also funding TSMC to move to Arizona. I don't got nothing against this, but domestic chip-makers should also get a piece of the pie, right?
--------------
In any case, this isn't "Pork". Pork is when you build something useless in somebody's state (aka: The Bridge to Nowhere), so that those "somebodies" vote on some other bill.
Aka: Build a bridge in Alaska to nowhere, and __suddenly__ your Alaskan senators / house members starts the process of log-rolling and helps out on other bills.
------
Pork was incredibly common in the 80s and 90s, and has been cut down in the modern era. Today, its impossible to get logrolling, because too many people have demonized pork.
If you want to get Senators / House members to support a large-scale bill, you HAVE to have something in there for them. Imagine how much easier it would have been to get Joe Manchin to vote for the infrastructure bill if we tossed in a few coal subsidies (aka: West Virginia Pork) into the infrastructure bill.
Alas: today's citizens are ignorant to the wonders of pork. Its actually very easy to convince other senators / house members to do something useful. You just gotta give them a bit of pork.
-------
Pork isn't about making the underlying thing more efficient. Its about convincing leaders to support the project to begin with and get it greenlit in the first place. Without pork, you can't effectively build a coalition. And today's Senate / House is proof of that, its incredibly difficult to get anything done because of our anti-Pork culture.
Life was better with pork. We did our grand experiment over the past 30 years by demonizing pork, and it failed. Time to bring Pork back.
I'm simply a pragmatist. We know Congress is perpetually deadlocked and that we haven't had a proper budget in damn near a decade. Maybe its time to whip out the old Pork technique. That's all I'm saying.
Anyone who studied how these budgets got passed 30 or 40 years ago knows: Pork was essential in building coalitions (especially across party lines). You scratch my back, I scratch yours. If you help me pass X Bill, I'll write a useless thing into the Bill that benefits your state.
Lets bring it back. Lets get Congressmen working together again.
> For national security, we need a state of affairs wherein it is profitable to build and run a plant in the U.S. that responds primarily to the dynamic needs of private actors.
National Security should revolve around the profit motives of a handful of private actors? Hard pass.
DOD is a rural jobs program. It functions as intended.
Other branches of the federal government that aren't designed to be jobs programs (e.g., NSF) do tend to make good investments and get things done.
We sit around and argue about this while TSMC was founded by the Taiwanese government. The Taiwanese gov't is likely still a giant shareholder.
I am curious as to what HN readers think is different about the two situations. As in, why can't it work in the USA?
Fun TSMC fact: Morris Chang was never given any options as part of his original pay package. He used his own money to buy shares.
Unfortunately we are still stuck with the legacy of Reaganism, which was essentially the undoing of the New Deal notion that the government could do good for the country and the average person.
The new deal involved governments giving billions in grants to multibillion dollar multinationals?
This was followed almost immediately by WWII, which was more of the same.
But then all the people making a living from this didn't want the money train to stop when the war ended so they hired lobbyists and the military industrial complex was born. Which made government ineffective, wasteful, and disdainful of the very people it was meant to serve. Enter Reagan.
To get back to the old thing you would have to first dissolve the military industrial complex. And then somehow prevent it from reforming.
The government already spends a higher percentage of GDP than they did during the 1930s, and real GDP per capita is also significantly higher so this is more even in real dollars per capita than it was during the New Deal. The problem is that parasites are sucking it all out and it doesn't go where you want it to go because of the corruption.
Under his administration, military spending (deficit spending, mind you, none of this was paid for) increased:
"Reagan significantly increased public expenditures, primarily the Department of Defense, which rose (in constant 2000 dollars) from $267.1 billion in 1980 (4.9% of GDP and 22.7% of public expenditure) to $393.1 billion in 1988 (5.8% of GDP and 27.3% of public expenditure); most of those years military spending was about 6% of GDP, exceeding this number in 4 different years. All these numbers had not been seen since the end of U.S. involvement in the Vietnam War in 1973."
There is no doubt that the New Deal platform was becoming somewhat wasteful toward the end of its' era, but it was still serving the broader public. Post-Reagan the government has been far less involved in directing the economy toward the benefit of the majority, and the conservative movement he fostered has worked tireless to ensure people don't trust the government to do anything.
Contemporary example: Instead of having the government build new housing developments and then sell them into the market, any it does build are designated as "affordable housing" with restricted access to limit their impact on the market price of housing, i.e. to keep them from actually making housing more affordable. Meanwhile there are rental subsidies that don't increase the housing stock and thereby primarily benefit landlords. How do you fix this when the problem is powerful lobbyists capable of unseating politicians who try?
Reagan never solved this. Government spending didn't really go down. The theory ("starve the beast") was that if they lowered taxes, the tax reduction would be sticky (because people like that) and then the budget crunch would cause politicians to have to trim the pork. But the cuts never happened. The lobbyists were too strong. So now we just run bigger deficits.
Nobody has ever solved the problem. If you give the modern US government a dollar, it goes to building tanks the army doesn't want or inflating the cost of healthcare, because those industries have strong lobbyists. To get it to go to building dams or meaningfully increasing the housing stock or anything like that, you not only have to fight Lockheed Martin for the dollar, you also have to fight the coal lobby who don't want more hydro, or the National Association of Realtors who don't want lower housing costs.
If you can't answer how to solve that, you can't get publicly beneficial government projects. And what we get instead is very expensive trash.
It would just seem kind of silly if the end result of subsidizing Intel's expansions was that Intel could outcompete other American companies (AMD & Nvidia) because of it.
For example, if the military must buy 50,000 of something it doesn't need from some military contractor located in some senator's district, that's a canonical example of pork.
The line here is subtle. The exact same purchase could be made through a competitive bidding process.
If I'm doing space research, that's probably good. If I mandate that NASA do a specific research project through a specific vendor, that's probably pork.
In most cases, pork isn't obvious. The bill won't say "NASA will give $50 million to Acme Inc," but rather "NASA will purchase $50 million worth of sheet metal from a veteran-owned company with at least 7 years experience in aluminum-matrix composite manufacturing, ..." and by the time you include all the requirements, what do you know, there's only one business which fits the bill.
Space Launch System - I think the funding line both required they use a bunch of tech (only made by certain vendors) but also prohibited any competitive bidding or evaluation of other approaches. $20B later we might get a rocket that costs $2B per launch? OUCH!
More common are the funny requirements docs. 5 years of experience but not more than 10 to reserve project for "new entrants". Located in a XXX district (improvement, benefit etc).
That was my point. A new train line in someone’s district; pork or infrastructure?
65 billion for “broadband” in language written by Comcast’s lobby machine. Pork or infrastructure?
From all appearances it depends on who is in power and if there’s enough pork to buy off a dozen congressmen on the other side of the isle, it’s no longer pork but “infrastructure.”
They could do the same thing with a competitive bidding system more cost effectively no?
Saying "They're already poor, let's not invest in their infrastructure" sounds like a self-fulfilling prophecy to me - they'll stay poor unless we invest in them?
But then the Senators from New York represent the finance industry and the Senators from Arkansas represent the forestry industry. So the Senators from New York make sure that nobody is ever allowed to disintermediate Visa and MasterCard and the Senators from Arkansas get big federal money for a lock and dam system.
Both of these are simple corruption.
No one would have built hydroelectric and electrified Appalachia, because why would you?
But once that investment was made, the cheap power did attract new industries, as now those regions had something others did not.
A foreign country (e.g. China) wants the industry to come there, so they provide subsidies. The industry moves, because hey, subsidies. Now the industry is there. Do you want them to come back?
The alternative to counter-subsidies is tariffs on countries that provide subsidies, but then people whinge about that too.
But some people whinge when anyone claims corporations should follow even the smallest accountability or regulation.
The Great American Giant (Intel) could never have existed without massive R&D spend in the post-war period, not to mention the fact that it more recently had its ass completely whipped by a government-funded company (the aptly-named Taiwan Semiconductor Manufacturing Company Limited).
But, anyways, the answer to your question is: because government ==== bad.
You can protect an industry without explicitly handing billions of dollars to one single player in a corrupt process that doesn’t actually help the company or the country be competitive.
However, Intel is not that much to blame here. GlobalFoundries bowed out of the node race on it's own.
And honestly, as the work of economists like Card/Krueger show - many of the so-called common sense economics understandings have very shaky evidence foundations.
Long story short, there is very little data in evidence of such claims.
I'm not so sure about that. People like to point out Johnson's wheeling and dealing but there was a broad consensus in US politics, he had a near mastery of the senate, and he ultimately broke the existing party alliances. Looking further back you really see most large bills getting done through essentially 1 party rule or during a crisis. FDR had gigantic majorities.
I'm not opposed to spending cash to make sure we have competitive semiconductor factories, but until very recently, Intel wasn't even that! They only produced their own chips, another massive 20-year industry shift that their executives somehow missed?
Where are the consequences for those long gone utter failures of leadership?
Intel has consistently increased their R&D spending for the past 5 years: https://finance.yahoo.com/quote/INTC/financials?p=INTC (see financials tab)
Yes you are wasting money on a company so that they are still selling something at the end of the day regardless of quality or sales . That is my point .
Specialization is more efficient in the short term. Free market pushes chip manufacturing to be a monopoly or duopoly. That's not resilient, though.
I don't mind my tax dollars subsidizing these industries remaining in the US, though. Indeed, I think it's good policy to have all major regions of the world be self-sustaining (e.g. not France, but the EU as a whole).
For example, the government can:
1. Provide market incentives. For example, the federal government can give a 25% subsidies for school Chromebooks/laptops/etc., require the CPUs and RAM to be manufactured in the US.
2. Provide guaranteed markets. The US military is an efficient place to do this.
3. Provide subsidies with outcomes. If Intel wastes the money, Intel stock moves from investors to taxpayers.
Personally, I'm also a big fan of tariffs. Free trade zones are great, but tariffs between free trade zones on essential goods can encourage local production. If US CPUs cost 25% more in the EU, and EU CPUs cost 25% more in the US, there's a strong incentive to have local manufacturing infrastructure.
Government has worked well before, many times in US history.
We're already competing against governments subsidizing this work. The Chinese, the Europeans, everyone else who is competitive is subsidizing their industries, to bring, keep, and grow the knowledge base locally.
To be against government spending here, is to hand all of semiconductor manufacturing over to other countries in the name of the "free market" - it is not free, and we've decided to "play ball" by subsidizing the industries here.
FINALLY.
Taiwan in particularly, the salaries these guys get and how many there are is rediculous (low salary, lots of folks in field).
Nine governors press U.S. lawmakers to pass data protection and privacy bills.
The two headlines that don't exist but which flash into my mind when I see something like this.
The key word in this title is not "bill", but rather "funding". That is, they could subsidize the industry in their own states without any involvement from U.S. lawmakers, but instead choose to ask for outside money.
> Hunt, who co-authored a report on the subject [bolstering the US semiconductor workforce at all levels] last year, noted that 40% of the domestic semiconductor workforce is foreign-born and that, despite the field’s growth, the number of U.S.-born graduate students in it has been stagnant for 30 years. He said the U.S. should admit more foreign workers with specialized skills to help satisfy the demand for talent that CHIPS Act funding could spur. In the longer term, he said the U.S. should build up domestic talent, such as through fellowships and ensuring university students have access to equipment used in industry.
https://www.aip.org/fyi/2021/house-hearing-airs-cases-and-ag...
The article touches a few times on the issue of staffing new domestic production facilities. I can't help but think that the prospect of staffing fabs, paid for in no small part by taxpayer dollars, with foreign workers is a recipe for blowback.
Now that we're so much more advanced it's hard to imagine how something like this could apply or work again. Google bankrolling some runs of 130nm chips is one of the rare few examples we have, of accessible means to re-learning how to make chips. But whatever it takes, I feel like the US must create not just chipmaking capabilities, but conditions for talent to arise & develop & grow; we must re-create the Silicon Foundry. We've got to put these capabilities into the hands of our people, people of all kinds.
My ideal bill includes provisions that these fabs have to provide some runs to small to entities/schools, and has to have like 1/3rd or 1/2 their output be open to whomever. If the goal is only to make more chips, the US is not going to create a sustainable chipmaking pipeline. It won't develop the talent & knowledge that is essential to being a capable, competent power at chipmaking.
Morris Chang, founder of TSMC, outlined four advantages Taiwan has. Talent was the top[1]. Taiwan is a country rich chipmaking knowledge, there's seemingly no pool of talent like it. The US can't simply fund our way through this problem, can't create effective competition just by throwing money at companies: we need to create the conditions for ongoing success. We need to create our own talent. Open fabs are needed to allow new companies & talents to grow. Giving students & learners access to chipmaking is required to create the lifelong dedication & pursuits that great great yields. Currently there is little access to the means to personal success in chipmaking, and the US needs to change that, if it want's to turn things around.
[1] https://interconnected.blog/morris-chang-global-semiconducto...
In the mid- to late-1970s (far before Korea entered the fray), Taiwan was the global epicenter of illegal clone manufacturing. Markets around the world - but especially Hong Kong - had a steady supply of cheap Taiwan-made Apple and (later) IBM clones.
The ironic part of the story is that cloning provided Taiwan with manufacturing expertise, and in short order, they were awarded legal contracts to manufacture legal computers from US companies for export to Europe and America. The Taiwanese government sent Taiwanese engineers to the US for training, and they returned to kickstart companies such as Acer, which originally started out manufacturing OEM clones for Canon and Siemens, but later started releasing computers under its own brand.
The conditions for incubating talent in Taiwan arose from piracy and a demand for cheap computers and parts - not sure if there's any modern equivalent to start the next wave of talent in Europe and North America.
Do these states (aside from California) have fabs, or rather any hope of getting new fabs? What's in it for them?
[1] https://en.wikipedia.org/wiki/List_of_semiconductor_fabricat...
Kentucky, Alabama, and Michigan are huge automobile assembly states [1] who are hurting from chip shortages. KY/AL are especially dependent on the auto assembly industry for non-ag employment.
California has Tesla, of course, but also likely to benefit from chip fabrication.
Wisconsin is kind of a historical auto state, so some solidarity makes sense. More importantly, though, I think WI+IL see potential for chip fab in their states because of their globally-unique fresh water source (Lake Michigan), and NC is probably on board because they see potential to benefit from domestic chip fabs in research triangle. Pennsylvania probably fits into this "one foot in because of auto suppliers and the other foot in because of tech" camp also.
The huge elephant in the room is that Ohio, also a huge auto assembly state, isn't on board. I can't really explain why Ohio's not on board, other than maybe their governor playing petty politics.
[1] https://en.wikipedia.org/wiki/List_of_automotive_assembly_pl...
How does it help to recruit specialists? How would it help them buy the manufacturing hardware which I heard is usually on 1 year+ lead times? What do you actually physically do with the money if you get it?
People like Intel have the cash to build a plant, if they can get hold of the hardware, they have almost guaranteed 100% utilisation as soon as the plant is open and will quickly recoup their costs to extend faciliites.
If they can't get the hardware and people to operate it, then the money seems useless?
> “The global auto chip shortage has hit Michigan and states across the country hard, idling plants and slowing production, threatening thousands of auto-related jobs up and down the supply chain,” Whitmer said in a statement. “With no end in sight, it’s clear we have no time to lose if we’re going to protect jobs and maintain our competitive edge."
Seems what they want to money for is not to build new hardware, to recruit new specialists or buy new plants, it would be to retain the current workforce and I guess provide funds for the plants while they recover from something. What they are recovering from is unclear.
next batch of MBAs, hey we are having too much expenses and issues with assembly and logistics, let's outsource more parts to China so we can assemble there cheaply and ship ready product to warehouses globally directly from China!
next batch of MBAs, hey why do we need those bunch of old grumpy engineers? let's decrease salaries and get rid half of them, we operate as a franchise/brand operator now.
next batch of MBAs, hey we are having high employee attrition rate of engineers in our Chinese facilities, let's increase their salary or outsource to Urumqi, doesn't matter if they are having human right issues, its cheap anyway.
CEO, Hey government we need bailout, we need your money, we are creating jobs (not here), we can't innovate fast enough (yeah sure, we don't have hands on talent anymore), China is leading now (yeah of course)
I wonder will China make same mistake in the future when outsourcing jobs to USA, thinking it has cheap labor?
This wasn't my understanding of what started the Phyisical Tech -> China exodus.
Growing up in the 90's, my understanding was we started outsourcing manufacturing to other countries to side-step export controls. In the U.S., you could export a research paper on encryption but couldn't export the implementation. So chip manufactures were faced with a decision: create two chips, one w/ encryption for domestic and one w/o encryption for foreign markets OR produce a single product overseas. I think they made the right decision, when faced with the crypto wars, but it knee-capped our domestic manufacturing and contributed to China's dominance in manufacturing.
An article going into the incentives created during this era by EFF founder John Perry: https://www.eff.org/pages/decrypting-puzzle-palace
I don’t think “crypto” was front of mind. Suspect it was “let’s build a computer” and the lawyers said “you can’t sell that in x market if you make it in the U.S.” so they didn’t.
Fair to say that second and third order effects of the legislation were the real focus for manufacturers?
It is going to get worse with the new stimulus bill as companies now must compete against the government for resources more than ever. You need manual labor? So do all the contractors doing govt jobs. You need engineers? So do all the companies with billion dollar energy contracts. Nothing wrong with this long term, it is going to be the sudden jolt that causes a lot of issues.
Yeah, somewhere along the line, we decided in America we can have profits without risks. As well, we believe you can make money with nothing more than money itself.
With all due respect, you don’t know what you’re talking about regarding this industry.