> Except tarrifs rarely help any of that: there's already extensive regulations in place to require local sourcing for defence critical components, all the way down the supply chain.
This is too limited in thinking. It's not just about "defence critical components", but the know-how and having the production workflow knowledge. It's all well and good to have rules on what goes into frigate, but if you don't have the shipyards to build things then it's a bit of a moot point:
* https://www.csis.org/analysis/ship-wars-confronting-chinas-d...
> You don't get a defence manufacturing industry without actually paying for a defence manufacturing industry.
It's not just about industry but about capacity as well: if you have (in this example) only (say) 4 shipyards you're going to have a tough time beating someone who has 40.
This is presuming you get to keep the shipyards. They're no use to you if they're all blown to hell in the first 48 hours - ships take months to build at minimum. If you lose your navy in the mean time, you won't be building anything.
This is the problem with these assumptions: they're all rooted in the industrial warfare the US won in WW2 but are not contextually accurate to today. WW2 wasn't fought with satellite targeting and precision cruise missiles which could be fired from half the planet away. Ukraine is currently hitting targets on the other side of Russia - "behind the lines" doesn't really exist for strategic assets anymore.
Giving up your economy for a future war with China that may or may not happen is frankly idiotic. The US already has thousands of nuclear warheads in storage so what are you afraid of? This is basically how the USSR collapsed.