[clarky07]

there is no data to validate either model. there are far far too many variables, and you can't test both of them side by side by definition. If one fails in a certain environment, you can't say the other would have succeeded. We've been doing keynes for quite awhile now, and it has mostly been going poorly. There is no way to say that doing the opposite would have gone better though. It might have, it might have gone the same, it might have done worse. We don't know, and we can't know, because we can't go back to 2008 and try it.

We could switch tomorrow, and again it wouldn't matter if it succeeded or failed. You couldn't say that doing the opposite would have a different total outcome.