One of my favourite economists blogging at the moment is Scott Sumner: he's not afraid to actually go out and ask what the real evidence for this or that theory is. Sorta the opposite of the archetypal macro-economist: yes, yes, that's all very well in practice, but does it work in theory?
Here he has a quite magnificent look at the pretentions of Keyneisan theory on the subject of fiscal stimulus. Once we've discarded those times when either monetary or fiscal policy could have reversed a recession or slump (for very much part of the canon is that when we're at the zero interest bound monetary policy cannot work and thus fiscal splurging is the only possible option) we're left with only three real life examples that we can work with:
1. The US in the 1930s
2. Japan post-1994
3. The period since September 2008
Those really are the only (modern) times when we've got both the slump and zero nominal interest rates. So those are the three times when fiscal expansion should have worked.
Yet as we know about the first two, they didn't/haven't. The US took a full decade to recover: the 1920/21 recession was near as deep but was over in a couple of years. Japan has been piling up the debt to where it is now nearly twice GDP and they've still not been able to trigger that desired faster growth.
Einstein's definition of madness was doing the same thing over and over again while expecting a different outcome. Is this what we're doing or do we think that economics comes with a special "third time lucky" offer?