To claim, however, that the evidence suggests it is working–and that we need more of it–is nonsense for two reasons. The first, which ought to be obvious, is that we only get one observation on events. To draw a causal connection between the stimulus and the fact that we haven’t plunged into another Great Depression seems bold, to say the least. Since we don’t have a parallel universe in which to play out events without the stimulus, we can’t refute it…

The other reason why it is illogical to claim a boost from the stimulus is that, for the most part, it hasn’t gone out the door yet…

[T]he most important stuff–the discretionary spending on infrastructure–has hardly started. By the end of the fiscal year, only 11% of the budgeted discretionary spending on highways, mass transit, energy efficiency and medical infrastructure will have gone out the door. This is the really direct government spending that many associate with the stimulus. By the end of fiscal 2010, Elmendorf estimates that only 47% of the infrastructure spending will have occurred.