Source item

RATIONAL INDUSTRIAL POLICY This is just theoretical at this point, but suppose a rational America wanted to wean itself off key Chinese exports in a surgical manner. What would that look like? (1) First, get data on current US supply chains. Because the state is ready to compel, it should ask for import, customer, and vendor lists on corporate tax returns. It would require paperwork, but you could partially automate it by allowing companies to link their bank transaction data to a government website, and it’s way cheaper than tariffs. (2) Equipped with the data on who each company is buying from and selling to, you get a visualization of the actual US supply chain in all its enormous complexity, like this but on a much larger scale: (3) Now you identify genuinely strategic areas. You do not care about French wine or Canadian maple syrup. You do care if the US military's vendor’s vendors are located in China, as per this Pentagon-commissioned study: (4) For each foreign vendor in those areas, you gather ALL the customers of that vendor and determine what the capital cost and time would be to stand up a domestic facility capable of producing the components they need. (5) Suppose for example they all need a special kind of widget, and they collectively buy $30M per year from this foreign vendor, and the factory to create that widget will cost $100M and take a year to set up. They now pool their capital (likely with government assistance) to set up a US-based alternative, and gradually shift over their spend once it goes online in a year. (6) This is what incentive alignment looks like. It is in the national interest and ALSO their business interest to collectively crowdfund a new supplier, because a second domestic alternative would put price pressure on the foreign supplier!Moreover, their joint investment has the prospect of venture returns, especially if they all shift over their business. It’s a public-private industry consortium, like Sematech in the 80s-90s, which is worth studying in detail (link in reply). (7) You parallelize this across many different foreign vendors, recognizing that some are too hard to tackle in your first pass. You use the money printer to strategically disburse cash whenever the customers can’t collectively afford the $100M factory. But whenever you do this you take a stake for the state, so the people see an ROI. (8) There are already state level funds like CalPERS or the UT endowment that could be repurposed to do this. Or you set up a sovereign wealth fund. The main point is to get proven engineers, entrepreneurs, and investors making these capital allocation decisions, even if they’re rewarded more with patriotic glory rather than money. (9) As you set them up, you recognize that not all of these new domestic suppliers will succeed, and moreover that not all of them will immediately succeed. It will take years and many of these consortia will fail. (10) But those that win allow you to then go one step up the value chain, moving key widget manufacturers onshore. You repeat this for the next set of foreign vendors, the ones you couldn’t tackle in the first place. (11) At every step you ask whether this is more of a “Canadian maple syrup on my pancakes” issue or a “Chinese parts in my missile” issue. You ignore the former because they’re military allies and friends, and focus entirely on the latter. Anyway. That’s how a rational country would do it. Get the data, focus on the most strategically important foreign suppliers, use public-private partnerships like Sematech to fund domestic alternatives, and then rinse and repeat as they succeed. REFERENCES (1) Here's a good piece from 2011 on Sematech, which is part of how the US responded to the Japanese challenge. Unfortunately, I don't think today's America has the patience for something like this, because they want a quick fix with a button press. But it is possible that some consortium of the EU, Canada, Australia, India, Japan, South Korea, Vietnam, and the like could do this to create second suppliers for key Chinese components: https://t.co/39wt9V0AIq (2) Regarding public-private partnerships, that is how the Chinese responded to the ASML sanctions. They set up a public-private partnership to collaborate on EUV lithography, they have a state-owned firm, they're funding multiple teams, and they're aligning incentives. https://t.co/Ix9owQb8mP (3) One potentially obvious point is that an international public-private consortium outside the US will need a neutral ground to hold the capitalization table and allow the global Internet to participate in it. There are various options for a new Switzerland that I won't get into here. (4) With that said, simply saying one will do a public-private partnership is like saying one will do venture capital. It is very easy to spend money and very hard to make quantifiable returns! There are many details in the execution. Much of the time, a public-private partnership just turns into an excuse for graft, like the green jobs boondoggle of the last admin. And even when it isn't graft, the best guys still can't always pull them off, because coordination is hard. (5) Still: given that the goal is to compete with an Asian giant, one should at least understand (a) how America responded to the last Asian giant (Japan) with a chip consortium like Sematech, and (b) how China is responding to ASML with their own public-private consortia for EUV, and (c) how one might use the Internet to coordinate similar consortia to get second suppliers as alternatives to key Chinese vendors.