Yes clearly they are, but few choose to break the numbers down to look at this.

On the extrapolated GDP numbers, forecasting is a less accurate science than weather forcasting. The degree of accuracy of GDP estimates more than 1 quarter forward isn’t exactly impressive. Normally we have to wait for all the revisions to even get accurate numbers for the past.

Incidentally, and talking of which, Spain last week made an upwards revision of its GDP numbers based for the last four years: based precisely on allowing for the number of undocumented workers active in the Spanish economy. I’ll probably post something on this during the week.

BTW: nice blog Jasper :).

And wouldn’t you expect that an extrapolation of GDP-numbers into the future already includes the benefits of current and future immigration?

Oh, of course. You’d have to do a real growth accounting exercise to see what were the real benefits. All I think Mandel is drawing attention to is that there is a substantial up-front plus element which normally isn’t included in the evaluation.

“is the cost of educating their children etc.”

Well children as children, I don’t see any special on-cost, as you have to do this with ‘natives’ too. If you mean ’special cultural adaptation costs’ for the education system etc, then obviously you are right.

On the other hand this too would have pluses and minuses, most of which would be hard to evaluate, which takes us back to the what you can and what you can’t measure problem.