I read through the comments on Tyler Cowen's recent stimulus posts and I notice the same phenomenon. I myself am an agnostic on the relative merits of spending vs. tax cuts, or permanent vs. temporary stimulus, because while I can tell convincing stories, the empirical evidence for the income shock we're currently seeking seems thin on the ground--there just aren't that many instances in history where conventional monetary policy has gotten to the "pushing on a string" point in a major industrial economy.
I'm less definitely a skeptic than Tyler--I'm more concerned by the composition of the stimulus than by its size. But Tyler's concerns are not unreasonable. Many of the concerns raised about the stimulus are not unreasonable. And the response to requests for better evidence are too often being met by enraged proponents metaphorically jumping up and down and screaming "Concede! Concede! Concede!" This is not usually the activity of someone who has solid empirical evidence and an irrefutable model backing him up. If the evidence is so overwhelming, why not just lay it out? What, exactly, is the model we're using; what are the assumptions about things like marginal propensity to consume; and what is the empirical evidence backing up these estimates? What's the justification, other than "it's a good way to fool the American people into supporting spending I want", for packaging so much permanent spending as stimulus, rather than debating those programs on their own merits?
We could use a lot more of these posts and articles, and a lot fewer "Here's another theoretical model under which the stimulus could be even more awesome than previously thought".
Incidentally, I think it's perfectly okay to say "there are a lot of unknowns here, but I still think it's worth trying". I'm with you, as far as the temporary components go. But I don't understand why no one has said this when the empirics just don't seem strong enough to support the confidence.
At least with the TARP, one could argue that action needed to be taken quickly--that the whole point was a fast firewall to block further damage. But most of the stimulus won't even be spent until after the 2009 fiscal year. What's the cost to taking a few weeks hashing things out? Is there something dangerous about making sure that the stimulus is going to be effective and well spent?
The one thing I can say is that the tendency of the proponents to get angry at anyone who questions them, and/or pick on side points, does not make me more confident in their judgement. It doesn't seem to me like the irritation of confidence. It seems like the anger of someone with something to hide.