Hoisted from Comments: Bruce Wilder on the weak link theory of economic development:
Some years ago, I was privy to the efforts of a company trying to build maritime shipping containers in China.
A shipping container is not a particularly sophisticated manufactured product -- more sophisticated than simple textiles, certainly, but most shipping containers include no motorized machinery. The most sophisticated aspects are the forged corner parts, which bear weight and must interlock with mounting hardware, and the requirements for dimensional consistency. The sides are pressed from rolled steel, and the floors are hardwood (at the time, teak was still used, when available), and the structure is riveted, bolted and welded together.
Despite rock-bottom wages, the container plant in China was consistently unprofitable. Why? Parts did not arrive on schedule. The quality of steel, forgings, other parts and finishes (paint) sourced locally was irregular. The local power grid was unreliable: power failures interrupted production and voltage spikes and brownouts damaged equipment. Local workers required a lot of training and supervision, and there was considerable turnover. Industrial services necessary to maintain the production equipment were difficult to obtain. Parts and repairs for the production equipment could be time-consuming to obtain. The purity of local water supplies impacted production quality. The list went on and on, with always the same result: finished output and productivity in the plant sucked big-time.
The venture changed hands and I lost track of them, but I would assume that successors eventually succeeded.
Such a concrete example may be instructive in reminding that high productivity is a result of getting very nearly everything right in a complex system.