> My uninformed speculation would be that the answer is something like 90% China for each hop along the chain, but those accumulate. 0.9^n becomes small with large enough n, so depending on the viewpoint it might be majority non-China…
If the cost allocated to China is 90% at each 'hop along the chain', then the total going to China across all hops is 90%, not 0.9^n. (This isn't like probability theory -- adding more links in the chain doesn't decrease the costs of earlier links.)
In any case, I have not found more recent estimates, but in 2018 one author estimated that only 3.5% of the manufacturing cost could be allocated to China. Their competitive advantage is in final assembly, but most of the pricey components are sourced elsewhere: https://theconversation.com/we-estimate-china-only-makes-8-4...
If the cost allocated to China is 90% at each 'hop along the chain', then the total going to China across all hops is 90%, not 0.9^n. (This isn't like probability theory -- adding more links in the chain doesn't decrease the costs of earlier links.)
In any case, I have not found more recent estimates, but in 2018 one author estimated that only 3.5% of the manufacturing cost could be allocated to China. Their competitive advantage is in final assembly, but most of the pricey components are sourced elsewhere: https://theconversation.com/we-estimate-china-only-makes-8-4...