Another way to say this is that the normally-presumed cause/effect relationship is actually backwards here. I.e. If you aggregate the outputs of industry over a long enough period, higher wages cause higher productivity and not visa versa.

This is a fun thought-experiment.