This thread is interesting, but it repeats some tired, Panglossian ideas about the causes of long-term deindustrialization. Deindustrialization isn't about steadily rising productivity growth rates, but rather declining rates of industrial output growth.
Here's an example of the pattern in France (used because the US numbers are skewed by bad reporting on productivity, as reported by Susan Houseman et. al.). Productivity growth rates are falling, but output growth rates are falling much more severely, leading to job loss. 
Declining output growth isn't due to a shift in demand to services, but rather a generalized slowdown in overall rates of economic expansion (GDP growth rates).
It didn't start with the rise of China, but rather the rise of Germany, Japan, etc. in earlier decades. The building out of global supply chains is a firm strategy undertaken in this hyper-competitive context of market overcrowded and consequent slowdowns in global MVA growth.
The notion that this is all a long term evolutionary feature of advanced economies is disproven by the extent to which deindustrialization is also occurring 'prematurely' in poorer countries, with much lower levels of GDP per capita.
Because it is associated with worsening economic stagnation, deindustrialization has not been a benign phenomenon. It is associated with rising employment insecurity and economic inequality.
The running down of the industrial economic engine is a major problem for capitalist societies. They are become increasingly incapable of neutralizing class conflict by ensuring that incomes rise broadly across society.
If you'd like the longer version of the argument, see my recent article in the New Left Review. newleftreview.org/issues/II119/a…
