Paul Krugman How late is too late?

January 27, 2009, 12:06 pm

One big question in the stimulus discussion is how much we should be worried about lags in implementation. Every economics textbook — mine too! — warns that stimulus based on public spending has a habit of peaking much too late, and therefore ends up being counterproductive. Now that we have a genuine CBO estimate of the timing of the stimulus plan, how worried should we be about the effects coming too late?

To answer this, it’s important to think about what the right criterion is.

It’s not a problem if some or even most of the stimulus arrives after the official recession, as determined by the NBER, is over. Why? Because in modern recessions, unemployment keeps rising long after the NBER has determined, based on things like industrial production, that the recession proper is over. You can see that the need for stimulus doesn’t end with the recession by the simple fact that in each of the last two recessions the Fed continued to cut interest rates long after the official cycle trough. if it’s good enough for the Fed, it’s good enough for fiscal policy.

So what is the right criterion? Actually, I think it’s quite straightforward. The reason we’re talking about fiscal policy is the fact that monetary policy is up against the zero lower bound. Stimulus will still be valuable as long as we’re still up against that bound — which is likely to be the case for a long time.

Again, I think it’s helpful to look at the last two recessions, even though we didn’t hit the zero bound either time. The 1990-1991 recession officially ended in March 1991; but the Fed kept cutting rates, and it didn’t start raising the target rate until Feb. 1992, 2 years and 11 months later. The 2001 recession officially ended in Nov. 2001; but the Fed kept cutting, and didn’t start raising rates until June 2004, 2 years and 7 months later. This suggests a long window, even after the recession officially ends, before the zero bound stops binding, and hence before the current strong case for fiscal expansion goes away.

Suppose, for example, that the recession ends this summer (which seems wildly optimistic). If recent experience is any guide, the Fed will still be keeping rates at zero 2 1/2 years later, that is, at the end of 2011.

Which brings me to the CBO report, which presents spending profiles in terms of fiscal years (which begin Oct. 1 of the previous calendar year). Given what I’ve said, any spending that comes in fiscal 2009, 2010, or 2011 is good, and it’s no tragedy if some of the spending trails off into the years following. CBO divides the plan into three parts: Division A, which is more or less stuff like infrastructure spending; Division B spending, which is stuff like increased food stamps;and Division B revenue, which is tax cuts.

By my count, 70 percent of the Division A stuff and 91 percent of the Division B spending comes within the fiscal 2009-2011 window. If you go up to the end of calendar 2011, we’re probably up to about 77 and 96 percent. That’s not at all bad.

So yes, lags in fiscal policy can be a problem — but they’re much less of a problem in the current context than the econ principles books might lead you to think.

Blogged with the Flock Browser