Will 3d printing and manufacturing advances (Foxconn can hire 3,000 employees overnight, but robotics lower labor needs) make the supply chain effect less important, lowering the costs of moving manufacturing back to the US? Will they come back to be closer to retailers?
The robot tech might be developed in the US (or where ever) but is still built in China. That makes sense, because it's used in China. And as expertise in building stuff concentrates in China all they need to do is keep quality high, keep costs low, and maybe keep an eye on human rights, and they'll keep the jobs.
The barriers to US companies become higher. You've got to get the machines. You then need people to run those machines. Those people will consider themselves to be skilled or semi-skilled, and require high wages. You need to run the machines 24hours, thus you need 3 shifts per day; two of those shifts will require higher wages than the day shift.
I doubt that huge volume manufacturing is going back to the US anytime soon.
The robot tech might be developed in the US (or where ever) but is still built in China. That makes sense, because it's used in China. And as expertise in building stuff concentrates in China all they need to do is keep quality high, keep costs low, and maybe keep an eye on human rights, and they'll keep the jobs.
If we assume raw materials are bought on the global market (so same price in the US and China), then the only cost variables are labor and shipping to market. Once Robotics drives the difference in labor costs below the difference in shipping costs robots in the US makes more sense than laborers in China. There is an inflection point even if you assume the use of robotics in China.
No benefits of scale for huge Chinese manufacturing ordering more raw materials?
No benefits from China holding large reserves of raw materials (eg rare earth metals)?
No benefits from Chinese dock workers getting paid a lot less than US dockworker? (See also transport from docks to factories; from factories to docks; from docks to markets.)
Just to give you one piece of data, we build the same product on both a fully automated line and a manual line. The manual line requires 22 people to run, the automated line requires 8. It doesn't take much time (12-18mo for the kind of product we're building there) to recoup the cost of the robotics.
However, there are some very good reasons for not automating everything. Manual intervention at the process level is extremely important for the NPI process and any time an ECO is being introduced for mass manufacturing. Generally speaking, too, you don't want to have a bunch of expensive capital equipment sitting idle just to handle occasional overflow, either.
Someone has to build and service the robots. They're also a lot harder to train than a human.
It could take three weeks of round-the-clock work to reprogram an assembly line of robots to perform a task a different way, to work out all the problems with a new process.
If you have people it only takes a ten minute demonstration and everyone's going to do it the new way.
Advances in machine vision have made the automated assembly line a lot better, but the tolerances are still extremely strict compared to people.
I think there's a better chance that retailers will move closer to the supply chain. Amazon doesn't really need to be based in the US, after all, and being closer to producers could have positive effects ( no uncertainty on release dates etc).
There would be positive effects from Amazon being closer to their suppliers, but there would be a huge negative being further from their customers. The main reason I order from Amazon is that everything gets to my door in 2 days, if they are shipping to me from China that value proposition is a lot harder to offer.
