I haven't stopped thinking about fiscal policy, but I'm not so keen to just repeat my previous posts (google "fiscal stimulus" and related phrases into the MR search function if you wish). I am, however, willing to listen to Arnold Kling:
From the Recalculation perspective, the economy needs to shift
resources out of some sectors and into others. The government is either
(a) permanently shifting resources from the private sector to
government or (b) temporarily shifting resources from the private
sector to government. If it is doing (a), then we are not facing mere
temporary deficits but permanent increases in government spending, and
eventually we will have to figure out how to pay for them. If it is
doing (b), then the Recalculation problem isn't really being solved.
Instead, at best the government is redistributing the pain from the
reallocation process out of the present and into the future. People who
otherwise would be unemployed can find temporary work on government
projects, but when those projects expire they will go back to being
unemployed. This is what makes the fiscal exit strategy so problematic.
In my view that is exactly correct. It's also why current gdp measures and projections aren't picking up how well the stimulus is or isn't working. A higher monetary velocity, as induced by fiscal policy, will indeed boost nominal gdp, but we need to see how well these projects stick, or not, with time.
Addendum: Matt Yglesias comments.