84 Comments

I’m all for reshoring and a strong industrial policy. Comparative advantage only makes sense in a perfect world. We don’t live there, so it’s best to leave simplistic models behind.

Expand full comment

Comparative advantage makes sense in the real world too, and many other imperfect worlds. It just doesn’t 100% determine everything.

Expand full comment
Jan 25Edited

These reindustrialization successes almost look too easy. At this rate, focused tariffs might be outright profitable. You just have to assume there are positive externalities to new business clusters in your country. Spillover benefits in new industries seem real enough - just look at Silicon Valley.

It's the same logic as the old "Dutch disease" or "resource curse", except America's distorting resource export isn't oil but Treasury bonds.

I think Michael Pettis' reasoning makes a lot more sense when you think about how resource exporters often underperform, and then apply the same logic to American exports of sovereign bonds.

We've been deindustrializing ourselves as a favor to China and the other surplus countries' employment policies. It's time to start growing our own industries again.

Expand full comment

I suspect the West writ large made the same mistake re Chinese manufactured goods as Germany made re Russian gas: believing that by offering itself as a large market it would make the exporter less willing to jeopardize that market by starting a major war.

It didn't work for early 20th-century Europe (where mutual economic independence failed to prevent World War I), it didn't dissuade Putin's Russia from imperialist revanchism, and it's unlikely to dissuade Xi's China from seeking to dominate the Pacific.

Expand full comment

Opening to China while keeping deficits = Σ(expenditures with NPV>0) makes sense regardless of the politics of the CCP. We just need to be MORE open to non-China.

Expand full comment

A key element will be how this helps the local-level American economy - manufacturing via automation and even AI can certainly help with speed, quality, and cost, but it certainly doesn't create the mass of jobs the "old school" manufacturing advocacy would have promised. If these are all small startups with regular people behind them reaping the benefits, then this creates bottoms up stimulus for the local economies. Unfortunately instead it will most likely be institutional investors reaping he windfalls, which may exacerbate income inequality.

Expand full comment

Providing liquidity for the global economy is a nice side hustle. It does not mean there should be a net capital inflow. Making deficits = Σ(expenditures with NPV>0) will not damage that line of work.

Expand full comment

I wonder how much that we could achieve the same outcome, at far less expense and deadweight loss to the US economy, by reducing the federal budget deficit and rebalancing the tax code away from consumption. Part of the reason that US manufacturing is uncompetitive is due to cheap imports because the dollar is strong due to capital inflows for US Treasuries. If domestic savings and investment were more balanced, would you see the same effect without having to play political favorites with industries? While I'm happy that US battery, semiconductor, and solar panel industries are doing well, is there a flip side to this where other kinds of manufacturing is down in industries that aren't politically favored?

Expand full comment

You can’t do what you’re saying while China, Germany, SK, and Japan are overproducing. If you cut the deficit meaningfully you’d spike unemployment and the dollar would still remain high relative to other countries. In fact I think you’d just see a worldwide depression. This is what I think trad economists either don’t understand or want to obscure. We have had an industrial policy for the last 20 years, it was just implemented by Beijing. If we continue to have open trade and capital markets and China keeps overproducing, there is no other option. It would be better for the US to lead a global effort to rebalance production and trade but I don’t see China going along with that.

Expand full comment

"We have had an industrial policy for the last 20 years, it was just implemented by Beijing."

My man, THAT is absolutely fucking brilliant!

Expand full comment

That line certainly got my attention. Noah's made the point before but this phrasing is wonderfully succinct.

Expand full comment

I'd believe what you were saying more if the US would ACTUALLY try it. In the last 40 years, the US has what ran a budget surplus for a mere 2 years? I wouldn't dismiss traditional economics if the US government would try it. Beijing doesn't control our politicians poor budget decisions in the federal government. If we've given them the noose to hang ourselves, that's our fault, not theirs.

Expand full comment

Yes we can!

Replace the wage tax with a VAT. Replace corporate and person income tax with a personal consumption tax so that in the aggregates state and federal deficits = Σ(expenditures with NPV>0)

Expand full comment

Excellent piece, Noah. But Big Oil is a top GOP constituency, so don't be surprised if Trump + GOP goes all-out to kill IRA. Solar & batteries are kryptonite to Fossil Fuels, Inc.

Expand full comment

The "industrial policy" that existed in my youth (born in the late 50's) on the SF Penisula was the huge Cold War spending on new electronic technology. This spending also impacted other places with high-end technology like Boston and even Portland OR (with Tektronix). It even existed in my graduate school years (mid 1980's) where in my group we had a separate lab funded by NRL (Naval Research Laboratory) to develop Li battery technology for US Naval use. But the Navy interest in Li battery technology in 1984 was entirely based on 1 time use in a torpedo or missile. This lab experienced a significant Li based fire, for an experiment in a good chemistry metal hood which blew out the plexiglass hood shield, enough to bring the fire department, and then shut down the lab for a few months.

The industrial policy we need must again be entirely focused on defense. It has worked before (WWII, Cold War) and it could work again if there was a long term commitment to buying domestic production of defense materials, but the political instability in the US might not allow for this. The DOD under one political cycle might get a commitment to buy at least $50B of domestic drone technology (batteries, motors, electronics, ...) a year but what business would commit to investing the massive capital required for this to produce a return without at least a decade of political stability.

Expand full comment

I’ve heard reports that US solar manufacturing is mostly just assembling Chinese components. Is this true or are we really manufacturing most of the value chain

Expand full comment

True... but that's the next step

Expand full comment

I was talking to a guy who was a senior engineer on the Nevada Panasonic gigafactory and he was taking about how China’s “light touch” environmental regulations made it so much easier to quickly iterate battery tech there.

America already has strengths in managerial skill and capital allocation, so if I were America I would focus on getting the cost of factor inputs down. So, land: cheap long term leases of federal land. Put it in West Texas or something. Energy: renewable energy superzones where you can get 80-90% of your power requirement from behind the meter and only need gas/grid for firming. No approvals required for generation or transmission. Labour: instant 5 year work permit for anyone with a job offer in the zone. The whole area is right-to-build zoned mixed use up to 6 stories so you can build accommodation on demand as much as possible. Certification: mini FDA, FAA, FERC and NRC to give fast track approvals for anything. Infrastructure: let the Spaniards and the Koreans bid for it. Get the Italians to manage the contractors.

You’d see some action then.

Expand full comment

Excellent points. Especially for projects distant from major metro areas.

My experience in Hangzhou in 2014, however, showed me that while the Chinese get infrastructure & manufacturing projects up and running very quickly, their urban planning is haphazard and generally sucks. Especially with regards to creating matching transportation infrastructure.

Expand full comment

File this under Things that Would Have Been Good To Publicize BEFORE the election.

Expand full comment

Actually, Noah did write about reindustrialization before the election, but not as clearly and forcefully as this particular post.

Expand full comment

Yes, I read those articles as well and, as you note, they were not as clear and forceful as this current piece. Not that it would have made a difference.

Expand full comment

And missed then as now the key elements:

a) state and federal deficits = Σ(expenditures with NPV>0)

b) cost benefit analysis of investments incentivized.

The idea that there were many high NPV investments just waiting for Joe Biden to break the "taboo" on "Industrial Policy" is almost ludicrous.

Expand full comment

Dear Mr. Hutchinson: I mean this in a kindly manner. Assertion A is written in math rather than English—you're going to lose most of the people who read these comments and 99% of politicians—the people who make decisions. Assertion B assumes that no cost benefit analysis was ever made. Assertion C is also an unsupported assertion.

Also, it is my understanding that many of these investments and subsidies were given for security reasons rather than purely economic. To reshore critical industries for defense.

Expand full comment

Thanks,

The "math speak" is to signal that this is incredibly detailed boring decisions weighing costs and benefits (pros and cons) about gazillions of issues. Lots of times I follow up with, "borrow only for investment." I'll try to do so more often. Of course my intended audience is not the politician but the assistant that has to explain things to the politician.

You have hit on one of the problems of discussing "Industrial Policy." For Hamilton (or the South Koreans) it was promoting things that woud just make the new country richer, which they could not do unless benefits exceeded costs, NPV>0. For others it is to achieve some non-commercial but economic end like reducing CO2 emissions to optimize the CO2 content of the atmosphere, "fight climate change." For others it is to avoid dependence on China for some critical import, "national security." Of course CO2 reduction and the non dependence on China has a benefit that should be compared to the costs of the investments. Like Burr, I wasn't in the room, but from the way these investments are discussed and defended, it does not sound like they were subjected to a cost benefit analysis. If I'm wrong (I have been challenged about the cost per unit of CO2 emitted by IRA projects) hopefully someone will tell me.

Expand full comment

Have you considered that using math speak might be construed as a message that you're not interested in communicating with those who do not grok mathspeak? I have never seen a mathematical formula in Paul Krugman, Peter Coy, or Noah Smith. I would not find your comical tag illuminating or communicative.

Expand full comment

I should be flattered to be compared with that group.

I have been thinking about trying to aim my posts more broadly, invest the effort in trying to change my style. You are not the first to point out my proclivity.

Expand full comment

The Solar Energy Industry Associatio lean claim that "American solar module factories can now produce enough to meet nearly all demand for solar in the U.S." is utterly fraudulent. The actual fact is that there is a the theoretical capacity to assemble enough *modules* in the US to meet current demand--not the actual cells that make up the modules. This is the equivalent of putting a case on a phone and claiming you manufactured the phone. The amount of cells being made in the US is still essentially zero. [1]

And note they talk about the theoretical manufacturing capacity--not how much is actually being manufactured. They don't even report that in the summary of the actual report. [2] I wonder why...

Sources:

[1] https://www.canarymedia.com/articles/solar/the-us-is-making-and-deploying-more-solar-panels-than-ever-before

[2] https://seia.org/research-resources/us-solar-market-insight/

Expand full comment

That is exactly right, and I mentioned it last time Noah touted this same SEIA report. Not only are there no factories making actual cells in the US, there are only two producers of polysilicon which is needed to make the cells in the US (Wacker Semi and Hemlock) but together the produce less than 10% of the top Chinese maker (Tongwei) and all top ten polysilicon producers are in China.

Expand full comment

Seems like you need to put on socks before you put on shoes. If we don’t have on-shore cell production, then it’s unlikely we’ll have on-shore module production that uses domestic cells. If we started with incentives for domestic cell manufacturing, who is going to assemble them into modules?

You need to go the other direction and start with modules (based on foreign-sourced cells) and then use predictable-in-advance industrial policy to force the new local module producers to use some percentage of locally-manufactured cells. This sends a clear signal that there will soon be a market for US-manufactured cells, driving investment to that area and so on.

I don’t know precisely what the IRA said about PV manufacturing as the decade went on, but I understand many of the EV incentives were designed to get stricter about battery production over time. First, local assembly of packs, then more requirements for domestic manufacturing of components, then more domestic sourcing of materials. All very crawl, walk, run.

Expand full comment

Ideally you would make the entire supply chain here, and start based on lead times. Polysilicon refinement is technically complex to produce¹ and takes more time to set up then cells. Cells need clean rooms and etching, but no expensive delicate ASML EUV machines. Module makers just do sheet metal, glass enclosures and wiring that can be done in a warehouse by minimally skilled workers. The modules have the least added value, and the cells and polysilicon can be exported and still retain added value here.

1. https://youtu.be/q6gBKTklJ18?si=HA3DeYBx1KRwJRzq. Polysilicon production by Hemlock.

Expand full comment

Again, this seems to be a ratification of what the IRA is doing?

As I understand, the goal of the IRA is to start by immediately funding many businesses in the supply chain. But the ones to have immediate real-world impacts will obviously be the ones that have the shortest infrastructure-construction lead times -- which as you point out, are the module makers (warehouses, minimally-skilled workers.) And since there is effectively no domestic cell production (yet), those industries will have to rely on foreign cells at first. But at the same time (I believe) the IRA conditions future-year subsidies on the increased use of domestic ingredients.

So module makers can start up immediately, thus creating a domestic industry to consume future domestic cell production. But then you announce that by year N1, you will condition future subsidies or credits for module makers on the use of (some) domestic cells. And by year N2 you'll tighten those conditions further, and so on. Meanwhile investors can take advantage of tax credits to build up the domestic cell industry, and they know they'll have a guaranteed domestic market in the future thanks to the requirements on module-makers and so on.

Expand full comment

Noah, I'd love to see an article focused on robotics and drones and our ability to manufacture those at scale? Given the impact of future wars and tech in these areas I'd love to know more how we are doing...

Expand full comment

Reindustrialization is possible? No way! xD

On a more serious note, I think the early (and still tentative) success of industrial policy suggests that a disturbingly-large amount of the macroeconomic assumptions girding current US policy are either ill-founded or need to be revisited. The idea that "manufacturing is dead" and "service-sector growth is the future" was literally beaten into my head in undergrad when I studied economics, but it never seemed to make sense to me. Moreover, recent success suggests that asymmetrical forces in international trade (the overvaluation of the dollar, overreliance on the finance sector, excessive consumption as opposed to national savings, etc.) were behind the "China shock" of the early 2000s, which in turn ridicules the notion that manufacturing job loss was due to purely structural forces such as "automation." Long story's short, I think this really damages the prestige and validity of mainstream macroeconomic theory.

Expand full comment

I'd be curious to see a deep dive (such as is possible) of the current state of the space race.

You allude to it here, but SpaceX can singlehandedly put more into orbit than every other company and government in the world combined. This has enormous implications for both the industrialization of space in the mid-long term as well as huge short-term ramifications for a global conflict where space dominance will be a key factor in determining the outcome.

Expand full comment

Don’t forget the big 3cm fab Utz is building with investment from Germany.

Expand full comment

Advanced packaging means 3cm chips can finally break the stranglehold that 5cm Pringles in a can have held for so long. Pringles were a great innovation in 1950 by Proctor & Gamble, and was then a great innovation over the Utz factories which could only produced 50 lbs of chips per hour and were packaged exclusively in low tech flimsy bags. Both are great American products and have survived for decades without industrial policy. Chinese imports of chips never took off, because Idaho and Maine dominate the raw material production.

Expand full comment

Ur, I do hope you mean "3nm" (3 nano meter) and not "3cm" (3 centi meter)!

Expand full comment

Also worth mentioning regarding semiconductor on-shoring: it can’t be done without immigration, period. A PhD is often needed, even for non-research manufacturing positions. I did a PhD at a top EE department in semiconductors and the majority of students (and faculty) were internationals, mostly from China, India, and Korea. Less appreciated is the fact that even among the Americans, naturalized Asians or their children made up at least half.

Not that I endorse Ramaswamy’s insults, but as an empirical matter, non-immigrant Americans are apparently unwilling to pick up that graduate textbook on QM or solid state physics.

Expand full comment

You exaggerate a bit and contradict yourself too. Are the children of naturalized Asians not also non-immigrant Americans?

You are not wrong that, historically, US students have not afforded as high prestige to STEM as students in other countries.

I believe this orientation is shifting though. When I was a student, ambitious kids wanted to pursue investment banking. Then 2008 happened. Nowadays, I think a lot more students are seeing the path to riches in STEM.

Expand full comment

The knock on effect of large manufacturing activities is extremely important. Having seen the rise and fall of the Polaroid corporation in the Boston area, Polaroid's demise left behind scores of small manufacturing activities - machine parts, plastics, chemicals .... Solar panel production and chip production will also be surrounded by a constellation of small supporting manufacturers.

Expand full comment

1. Trump thinks tariffs are the only tool

2. But even if you de incentivize foreign purchase you still need to have game changing domestic production

3. It has to be a coordinated effort or punishment and investment

- it’s the only way this works

Expand full comment