There is a quite interesting article in today's Guardian by Will Davies about, How Statistics Lost their Power. No it is nothing to do with the probability of rejecting the null when the alternative is true, but a rather fanciful discussion of the alleged decline of belief by mass publics in quantitative descriptions of their societies.
To my mind he mixes a lot of things up in the first section where the story of political arithmetic, trade indicators and the beginnings of the sample survey are all jumbled together. By the way he seems to bring the latter forward to the 1920s. The big innovations actually happened about 20 years earlier and were associated with people like the Norwegian Kier, but then again being a proponent of elitist quantification I guess that is just the sort of petty minded concern with accuracy that you would expect from me.
What he neglects to say anything about is how social scientists have themselves contributed to a generalized distrust not just in the use of numbers to describe the social world but even in their very calculation. Things were already bad in the 1970s, so bad that a prominent British Althusserian Marxist found it necessary to write a little book defending the use of official statistics! Well, if you were a Marxist you couldn't really admit it was all a social construction, after all the Master himself used the Blue Books pretty uncritically.
But after that it was pretty much downhill following the rise of autoethnography, poetry writing, epistemic communities etc.
The brave Marxist, after writing an auto-critique, took himself off to Australia.