Paul Krugman writes (citing Noah Smith) that he agrees with the empirical findings in my critique of the revival of Keynesian activism in the 2000s (the stimulus packages of 2001, 2008 and 2009). In particular, he writes that “it’s far from clear that the ARRA actually led to much of a rise in government spending, while the tax cuts that made up much of the stimulus were probably largely saved.”
But he then goes on to say that the stimulus was too small. That’s not what I found in my paper. As I stated in the paper, my “results do not lend support to” the view “that the stimulus was too small.” Rather the paper showed that “a larger stimulus package—with the proportions going to state and local grants, federal purchases, and transfers to individual the same as in ARRA—would show little change in government purchases or consumption.”
Now, I know that Krugman is trying to distinguish between good and bad Keynesian stimulus packages, and that he would like a stimulus package with higher proportions going to federal, state, and local government purchases than the 2009 stimulus, or, for that matter, the 2008 stimulus or the 2001 stimulus. But experiences from the 1970s raise serious doubts about the political and operational feasibility of such discretionary fiscal policy. So do recent experiences in many other countries, as shown by Hyun Seung Oh and Ricardo Reis.
In a simple Keynesian model, all the government has to do to combat a recession is quickly increase government purchases, but the difficulty with doing so in practice is one of the classic arguments against discretionary fiscal policy. Of course, it is not the only argument. Small or unreliable multipliers, the legacy of increased debt, the unpredictability and temporariness of such policies are some of the other arguments. Using dynamic models with expectations and incentives, I have found very small multipliers (around .5)
For these reasons I argued in the November 2008 article which Krugman cites that a better fiscal policy would be to rely on the automatic stabilizers and enact more permanent reductions in tax rates (or at least pledge not to increase tax rates in a recession).
As early as the summer of 2009 it was clear that ARRA was not working as intended, as John Cogan, Volker Wieland and I reported. Research since then has uncovered the reasons why. One reason is that very large stimulus grants to the states did not go to infrastructure spending as intended, and that’s what Ned Gramlich found out about Keynesian stimulus packages thirty years ago.