Mario Rizzo makes an important point. If, like Olivier Blanchard, you think the economic problem is pervasive “Knightian uncertainty” (i.e., folks don't even know how to construct the relevant probability distributions, and so can't begin to think about the expected return of various economic bets), then how is, say, federal fiscal policy supposed to help?
TARP money was going to be spent this way, and then that way, and then this way again. I was pretty sure the so-called “stimulus” package was going to pass the Senate, but then it didn't! Now they're changing the composition of spending and tax cuts. Once a bill is passed, how am I supposed to know it's going to be reliably implemented? Maybe things get hung up because every single member of the Obama administration turns out to be a tax-evader and/or lobbyist. I mean, it looks to me like politics is pretty chaotic.
I know this fact has got to be supremely annoying if you're a very confident technocratic macroeconomist type who knows what to do to abolish Knightian uncertainty. But it turns out that even the brightest and best of the best and the brightest, like Blanchard, are mere sideshows in the colorful, helter skelter carnival of American democracy. Precisely calculated monetary targets can be established and then ignored. Rigorous cost-benefit analysis may indicate the wisdom of a new bridge, which may be promised, and then left unbuilt. I realize that it is not helpful to sow uncertainty when we all need to take a deep breath, get in line behind the big boys, and calmly follow their advice. But it's also not clear to me why the big boys think government economic policy is to uncertain souls what Baby Gold Bond is to diaper rash.