I am a fan of the way interdependency makes war between great powers illogical — because the supply chain for microchips are many, many countries long. If you’re China, invading Taiwan doesn’t get you anything. It gets you control of some chip fabs, but that’s just one node in that whole thing. By the same token, it prevents wars over resources like oil.
On the flip side, we have strategic reserves of petroleum. And now, if you look at Intel and various lobbying bodies, we should have the equivalent for microchips… Intel wants like $90 billion in subsidies to build chip fabs in the US. The problem is, as we’ve seen through trade wars with China who has invested incredibly heavily in their own domestic chip fabrication capacity, that doesn’t matter a bit if you can’t get one machine from the second largest Dutch company ASML which makes the extremely specialized — I’ll just call it a laser-thingy to gloss over what it does, which etches these tiny features in chips.
When Intel and other lobbying bodies say, “Let’s build more chip fabs in the US, it will make us more resilient and prevent future chip shortages,” my response is “Maybe, but their own research has identified 50 different points in the global supply chain for chips where one geographic region controls more than 65 percent of the market.” To translate that, it seems to me there are 50 different single points of failure where, if a country shuts down, there’s a new pandemic, god forbid a tsunami or volcano or war, that single point of failure could shut down most of global chip production.
Unless we’re going to relocalize that entire supply chain — which doesn’t seem to be on the table — I’m not sure you can really can create that kind of strategic reserve of microchips. You can create more capacity. TSMC in Taiwan is doing that, and they’re building plants in Arizona and Japan. The market is doing its work. Company logisticians are realizing “we can’t have a single source for this.”
Nike is doing the same thing. Doesn’t matter if it’s chips. Someone told me on background, “name a giant global apparel manufacturer — they have set up carbon copies of their factories in China in Vietnam and places like that before the pandemic because they were worried about a trade war or a shooting war in the South China Sea.
That was Christopher Mims on the Realignment podcast: Supply Chain Crisis.
One nuance is that while DoD can have strategic reserves of unchanging commodities like JP-8 jet fuel, a strategic reserve (i.e., stockpile) of microchips doesn’t get you much. The chips you stockpiled last decade aren’t really much good under technology change. Of course, I should bite my tongue there because most of DoD’s systems are between 10 and 50 years old. But if DoD wants to scale the newest technologies to meet a near peer adversary, then it needs a strategic reserve of state-of-the-art manufacturing capacity (which is what I think Chris Mims was referring to).
I don’t think Mims really solved anything for my mental model of the supply chain problem. Globalization of supply chains won’t necessarily stop China from pursuing strategic ambitions. One person can make decisions regardless of the “logic” or consequences. And in any case, as Mims discussed, companies are actively de-coupling from China on their own. Its very hard to get a clear picture of the implications.
The fundamental tension in my mind is this: (1) There is mountains of evidence that relatively free markets generate lower costs and superior technologies; (2) relatively free markets are agnostic to geopolitical boundaries creating a complex web of interdependencies, and so when existential crises hit massive disruptions can occur.
Ultimately, the disruptions happening in the US in late 2021 are tame compared to what we should expect in a near-peer conflict. Coronavirus is not an innovating enemy that sinks ships and launches missiles. The US clearly needs its own capacity to carry out a fight. But the strength of the US is also in its international relationships and market economy. The US cannot solve this problem with central planning or mandates — it cannot beat China at its own game.
… whatever amount of funds should be reprioritized for mobilization could go directly into R&D for next generation manufacturing rather than “slack” in today’s manufacturing. If a conflict with China or Russia does accelerate first in the gray or proxy zones, then there will be some years lead time (more like a WWII than a WWI).
And another critical element is reforming the budget system, which is specifically set up to fund weapon end-items only and not enabling inputs like critical suppliers, machine tooling, facilities, workforce, etc. It presumes that if a fighter aircraft is procured, the supply chain will figure itself out. Perhaps that’s reasonable from a peacetime economic point of view, but DoD exists for wartime.
Leave a Reply