Jack Strocchi sent me a link to this piece by Karl Zinsmeister on America and the EU a week or so ago. I read as far as this para:

America’s poorest sub-groups, like African Americans, now have higher average income levels than the typical European.

What’s behind this? For one thing, Americans work harder: 72 percent of the U.S. population is at work, compared to only 58 percent in the E.U. American workers also put in more hours. And U.S. workers are more productive–an E.U. worker currently produces 73 cents worth of output in the same period of time a U.S. worker creates a dollar’s worth.

and concluded that the piece wasn’t worthy of serious attention. As I pointed out, multiplying these figures out (including the hours difference) implies that income per capita in the EU is less than half that in the US which is nonsense.

Jack did the sums and got Y = a(0.73) X b(0.80) X c(0.80)

Y = EU/US pc income ratio
a = EU/US labour productivity ratio (EU .73/US 1.0)
b = EU/US labour participation rate ratio (EU 0.58/US 0.72)
c = EU/US labour hours ratio (EU 1600pa/US 2000pa)

So Y = 0.4672
The correct figure, as reported by the US Bureau of Labor Statistics is around 65 per cent for the EU as a whole and 75 per cent for the leading EU nations like France and Germany, as well as Canada (not usually claimed to be in a state of irreparable decline). The big error in Zinsmeister’s piece is the number on output per hour worked – the US actually lags a number of European countries on this measure.

I’m pointing this out now, because the Zinsmeister piece has been taken up by a number of US bloggers, notably Steven Den Beste and Eric Raymond. Before accepting a statistical claim that seems ‘too good to be true’, it’s always a good idea to do a ‘back-of-the-envelope’ credibility check.

Update I’ve been in a bit of a rush today, so apologies for not mentioning that these links are via Bargarz, who has also posted on this and on Kupchan.

I should also note that the big economic difference between the US and the other leading countries is in hours worked per employee (about 20 percentage points of the 25 per cent difference between the US and the other leaders). The US is above average in both employment/population ratios and output per hour, but not the leader on either measure. The Scandavians lead on E/P (all that child care) and the Belgians on output per hour (they don’t put in a lot of hours, which makes it easier.

I stopped reading the Zinsmeister piece when I reached the first major error, but writers in the comments thread point out other unbelievable claims. I’ll try to do a more complete statement later.