I take the point that there are different ways of looking at the
labour reduction problem. It could be the costs of training that drives
things (the Hilferding approach) but this is not very practical for
empirical work. The approach which takes wages as the indicator of
labour quality is tractable, since wages are observed in various types of
data. Both approaches can be defended on textual and theoretical grounds
but one of them is more suitable for empirical research. Correct me
if there are studies which have modelled the Hilferding approach
empirically.
In solidarity,
Andrew Trigg.