The "stimulus wars" have recurred lately in the blogosphere, and I think they've frustratingly failed to focus on the central issues. In my mind these are:
1. In the United States of America, discretionary fiscal policy was not in fact effectively deployed to prevent a prolonged period of mass unemployment.
2. In the United States of America, we have experienced a tragically long period of tragically high unemployment that is doing unmistakable long-term damage to the country's productive capacity.
So the question is what kind of measures can be put into place to prevent this from recurring.
I think we need to accept that the United States is afflicted by a serious and persistent disagreement about the appropriate size, scope, and role of the federal government. Conservatives believe that interest group politics and voter myopia systematically push levels of public spending up to inappropriately high levels, while liberals believe (correctly this time) that interest group politics and voter myopia systematically push levels of publis spending down to inappropriately low levels. Consequently, any major discretionary fiscal policy moves are processed by the political system as a move in the endless game of long-run political conflict rather than moves in the game of short-term macroeconomic stabilization policy. For a country with the institutional set-up of a UK, Canada, or Sweden that's not necessarily a barrier. For a country with the political consensus of a Germany, it's not necessarily a barrier either. But given the institutional and ideological dynamics of the United States it is very unlikely that large demand shocks are going to be effectively managed by large discretionary fiscal stimulus programs.
So what can we do? I see three highly desirable policy steps:
1. We should aim for a long-term inflation rate of four or even five percent so that the Federal Reserve is much less likely to hit the "zero bound" and lose confidence in its own ability to shape the economy-wide demand picture.
2. We should make specific statutory provision for Fed injection of "helicopter money" into the economy. The metaphysics of fiscal vs monetary policy are less important than the fact that the Fed has the right institutional setup to conduct a joint fiscal-monetary action when needed. A Fed that can order money-financed payroll tax cuts that have zero impact on the deficit is never going to "run out of ammunition" in the war on demand shortfalls.
3. We should beef up automatic stabilizers in the budget by creating some kind of national rainy day fund that automatically releases unrestricted funds to state governments in times of recession. Some elected officials will use the money to avoid pro-cyclical service cuts and furloughs, while others will use it to finance tax cuts and we'll just live with disagreement about the best way to proceed.
As far as the 2009 ARRA, the best thing I've seen on this was a crude poll of economists that was presented at the American Economic Association meeting. On the question that's dominated partisan debate, there was overwhelmingly consensus that ARRA created jobs and made the unemployment rate lower than it otherwise would have been. But on the question of whether the overall costs of ARRA exceeded the overall benefits, respondents were much more sharply divided. Which is to say that a lot of economists, like a lot of Americans, have an overall view that government spending (including lots of spending that clearly supports employment) is generally wasteful and thus that a big stimulus bill is just part of the general waste trend and all things considered a bad idea. I disagree with that, and I don't regret that congress enacted the law but I think it's a fantasy to think that re-running the ARRA play is going to be the solution to future recessions.
In general, not enough people recognize that an aging society running a two percent inflation target is going to hit the zero bound quite frequently over the years and really needs to make concrete workable plans to deal with it. Most years, DC gets by without a single major snowstorm. But major snowstorms come frequently enough that the city needs the capacity to plow the streets without it being some huge weird panic situation with a lot of ad hoc scrambling around.