The better strategy is what everyone else has been doing, letting the third world catch up in labor costs for a few generations while ensuring you maintain the design capability. This does more than any ngo ever to decrease poverty and suffering, it will mean at some point in the future your labor won't be so starkly uncompetitive so eventually you can bring some manufacturing back, and it makes it less likely that the best and brightest arent' wasted on subsistence farming/made less due to childhood malnutrition in the third world, which should help us innovate faster in the future.
I suspect if manufacturing ever did come back it would be automated manufacturing.
None of the options are good for overpriced manufacturing labor in the first world that doesn't want to retrain/can't retrain.
Most of the options are good for the whole world, because manufacturing with less expensive labor usually translate to higher output for lower cost and the labor inflation that causes in the third world is in the really nasty subsistence level of the human suffering index rather than the can't afford a second car and/or the new video game console area of the human suffering index. The benefit to the world of labor price inflation in that level of the curve is so high it's not really comparable.
From the perspective of specific first world countries whether it is good or not depends on a mix of priorities from maintaining output capabilities in the face of potential war, maximizing the stuff your populace can get (bread and circuses limit political risk to rulers) and ensuring the spread between the upper and lower ends of society doesn't get so wide that the lower end looks up from its bread and circuses to revolt. Overall it will be a good strategy to move a significant portion elsewhere from this perspective as well.
Of course automation levels will continue to increase as they always have since the Industrial Revolution.
They exist in lower income countries, but the work quality is often shoddy and poor at best. The talented ones have already migrated to the West. Case in point, a factory we built for a project in the Middle East with a fully European team took just 9 months to finish from the start of laying the foundation stone. Similar projects in the same country by a Chinese or local company would have taken 3-4 years. Another American team executed a project in Singapore in 2.5 years, but that same project was estimated to take 6 years in India.
It's a sad waste of human potential due to toxic politics and unfavorable geography.
The people who say these kinds of things are usually pathetic specialists with a sole focus on their singular task, without looking at the big picture. They have limited interaction with the rest of the team/company/world, and think too highly of their talents and their jobs.
Tell a 50-yr old Java engineer to retrain in Python and Machine Learning, while juggling family, kids and a mortgage, and see how that goes. Newsflash, it doesn't go well. Heck, I've grown up on Android my entire student life, that now I can barely relearn mobile usage with an iPhone. One of the managing partners at my old firm, ex-Harvard and Wharton grad, stuck to using a Blackberry till 2020ish, because he couldn't figure out both.
Now expect a miner to learn how to code, like Joe Biden tells them to....
What is more insane is, that whole progress I see in IT is just joke. All the new tooling that supposed to be better, easier to use and more productive is actually more bloated, slower and works against you. Yeah, indeed, sometimes its easier to start with a tool (aka, clickty clickty) but thats all.
Im heavy CLI user and I follow UNIX philosophy. I love it, write many simple tools and use them together. It works so well. Why abandom it?
If you don't have 3+ years experience in AWS, Kubernetes, specific frameworks with knowledge about system design, your employment prospects are in the toilet right now. Nobody values generalists anymore.
Too often I see people say to the disappearance of manual labor jobs: good, retrain people to use computers. I have slowly come to the belief that there are some people for whom working with their hands is the only kind of sane work they can do.
You can call them blue-collar, or working class (these are the people incidentally that raised me, the people in my neighborhood growing up).
Why would anyone not want to work in a clean office with only a compiler as your boss and a big paycheck waiting for you every two weeks? When I was younger I assumed that anyone not in the tech field were people not smart enough to pass Algebra in high school. It had not occurred to me then that it might go beyond aptitude and instead be due to something more fundamental about a person's nature or character.
I know I'm getting hand-wavy with nature, character but I'm not sure what other words to use. I came to believe that though many people who are like me find some satisfaction from having solved a difficult problem, or having created a functional framework that can now do work to solve problems there are many others where satisfaction primarily comes from having changed something with their hands.
I don't know if all of that makes sense or resonates with anyone. I'll add this though, the older I get, the more I recognize even in myself the satisfaction that comes from doing something physical (even mowing the lawn as an example). I seem to take some pleasure from taking something unordered or in a raw state, and ending up with a thing in an ordered or "finished" state. It's no doubt why I have pursued woodworking as a hobby, print making, creating PCBs. (And that fresh mown lawn looks so much nicer than the ragged lawn that was there before.)
In any event, following that line of reasoning has made me realize we should keep a large number of manufacturing, construction, repair and other manual labor jobs so that we can serve the whole of society. Just expecting everyone to learn to code is misguided and will fail so many people.
It is sad to me that Germany appears to have tried this model, tried to keep manufacturing in their country, and yet it is purportedly failing. I'm not too optimistic about our future when we apparently measure progress by "cost of living standards".