Which goes to show, you probably can't measure it. Or is you can, you can choose what fits.
It's more a case of measuring different things (often
very different things) and giving them all the same name. Though the second bit ("...you can choose what fits...") is often true as well. And often the period being measured matters as well. BBC did a little graph of the OECD's figures for the UK;
You can interpret the figures represented there in all sorts of different ways.
Yes, but who's anecdotes? That's always the problem with measures of poverty (or whatever) that aren't based on very crude calculations (and the problems of very crude calculations are too obvious to have to point out...)