It’s US-only and doesn’t break down by income groups though – to get to the kind of level I’m looking for I’d need that one, plus similar sources or my own surveys covering at least four or five other countries; then I’d need to spend several weeks building a model that related consumption to social characteristics to apply to the countries where no data was available and to relate class/age/race stats to income stats.

Not so impressed by the AP article – see comments on your blog.

And it’s time for your weekly apology again.

He just won the John Bates Clark medal for best economist under 40, so presumably he knows his onions.