Problematic indeed, I can't possibly prove my premise, and therefore I must resort to a shabby partners in guilt argument. The gist of this is that my economic argument suffers from lack of definitive evidence to the same extent that all possible economic arguments must - basically they discuss abstract phenomena in empirical terms, and the data they can possibly take into account is invariably insufficient to properly justify any such argument. Economists can only ever get strong data by narrowing the focus of their inquiry, and therefore a truly justified theory explaining what pay rises are appropriate across a general population is logically impossible. Which leaves us with only the choice of whether to reject all talk of such things, or else to accept that there are serious epistemological limits to any conclusions we may arrive at.
I'm really perfectly happy either way; I have, as you can see, a very low opinion of any claims to scientific rigour and predictive certainty that any economist might make. Which is why I consider the report that ignited this thread to be little better than astrology.
To put it more bluntly still; I'm happy for the bathwater to go, and willing to throw the baby after it.