But stuff such as the price of staples in different localities is among the things that PPP can and does factor in. It can't, by definition, take account of the "grey/black" economy which might push the PPP down through its impact on prices -- prices which don't quite make sense, looking only at official GDP.
Whatever, surely there isn't one single measure for inter-country comparisons. PPP probably works well enough, give or take, for actual in-country living standards in some dimensions, better for some segments of the population than others. In an earlier post, Doug suggested that PPP might have the "fortunate side-effect" of making the poor seem less poor. But it might actually have the side-effect of making the middle income in the non-US/EU seem richer -- in so far as the basket used to estimate PPP might reflect a greater share of in-country production (or inputs for final consumption goods) than would be true of the consumption basket of the middle income. An instance would be the greater share of wheat-based products (inputs purchased at exchange rates) in the consumption of the middle income in tropical countries.
Exchange rate comparisons would be better if the idea is to get at the structure of global inequality; and to the degree that we come to have a single global market for goods and services, exchange rate comparisons do an increasingly better job.
kj khoo