Why manufacturing?

This came up as a tangent to an Erik Lund history-of-technology post; why worry about manufacturing?  Why is rearranging dirt more important to the economy than rearranging schedules?

(Anybody who thinks rearranging schedules is easy needs to try it with several competing sets of interests and priorities; for difficulty, the schedules might be more fundamentally so.)

Right now, we're in the early stages of the end of the world, in at least the sense that fall of the Western Roman Empire was the end of the world.

One of the things that means is that while capital creates long supply chains to maximize returns on investment, instability shortens your supply chains.  There's three ways this is certain to happen in the next five, and the next twenty, years.  ("Short term", if I don't watch what I'm saying.)

1. We're going off oil.  We're going off fossil carbon generally.  This is going to be a combination of solar PV being flat out less expensive, dawning horror at just how bad the climate swing is going to be (see if you can get housing sale data for Florida by age and education...), and the collapse of American hegemony ("the Oil Empire maintains the stable oil markets necessary to the deployment of all that capital" and "the Oil Empire manages to keep people from effectively opposing technologies like fracking" both apply).  Does this shift the basis of naval power?  That happened in the 1950 with nuclear submarines.  It may force notice; it is very likely going to break the commercial arrangements which came in with DREADNOUGHT's continuation of the Anglo thalassocracy.
2. Food security is going away.  Direct habitability is going away for at least places with large populations.  (Second order, "I can get food there", is a different and more general question.) This is going to do in economic predictability; the supply chains don't notice the people suffering, but do notice the lack of spending, the loss of productivity, and the shift in economic focus from what people want to what people need.  Also the inoperable ports and rail links.
3. We're at the front edge of an age of miracles so far as capability goes; there's clinical trials of effective anti-aging treatments reported, there's meta-materials, someone will eventually figure out how to get bulk graphene.  Normally innovation is irrelevant because incumbents can keep it from getting enough capital to threaten their markets.  The historical counter examples are all either extremely prolonged (see "clocks, Europe, 1400-1800") or side effects of three enormous wars.  (Napoleonic, Great War, Hitler's/Great Patriotic/Great Pacific; incumbents willing to be less rich in preference to maybe losing in a war of annihilation.)

ALL of these things will shorten the possible supply chains.  Not because anyone wants there to be shorter supply chains and a less capable economy, particularly, but because the predictability necessary to the present arrangement isn't going to hold.  (Has stopped holding, and isn't likely to come back.  Even if everybody with significant political power does the optimal thing for the next century over the whole planet, it's not likely to come back.)

So the answer for "why is manufacturing important?" is "shorter supply chains are less capable but more resilient; to be reasonably sure you'll have all the critical stuff, you need the ability to make it locally, for someone value of locally, and remember that value of locally is significantly political".

One awkward part is that VLSI is really difficult and takes enormous supply chains.  Communications isn't a luxury, and one thing that's likely to be interesting about the next twenty years is the difficulty of keeping the incumbent communications infrastructure working as the necessary supply chain length gets unmaintainable.  Here's hoping some equivalent to direct electron lithography gets established as a viable alternative Real Soon Now.

Another awkward part is no one knows what's going on, supply-chain wise; heat pumps and window glass and autoclaves and refrigeration and tractors aren't optional as functions, and doing it with electricity and aluminium and glass instead of coal and iron and brass is obviously entirely possible but making sure all the parts of the supply chain are functioning and close enough hasn't been addressed.  It ought to be, and it ought to be with some urgency.

(I am, tangentially, all for trade and for long-distance trade; we're desperately going to need it to be a bunch of local economies rather than one global one, though, because we're not going to be able to keep a global-supply-chains one running consistently this coming century.)

So -- gotta be able to make knives and abrasives and paint and bandages and soup pots and dust masks and shoes and socks (keep thinking about just how much stuff is on the real list; artificial light should come to mind) in a couple local places by a diversity of means, or there's no resilience of supply, and we need the resiliency of supply because there's going to be no meaningful agricultural or climatic predictability for at least the next hundred years.