Guesstimating the Stimulus

Although the details are still in flux, Christina Romer and Jared Bernstein (incoming CEA chair and chief economist for VP-elect Biden, respectively), have put out a rough estimate of the effects of the Obama administration stimulus package.Their estimates are based on a plan of $775 billion, spread over two years, including both spending and tax cuts. Paul Krugman says:

Kudos, by the way, to the administration-in-waiting for providing this — it will be a joy to argue policy with an administration that provides comprehensible, honest reports, not case studies in how to lie with statistics.

But the stimulus is not enough:

So this looks like an estimate from the Obama team itself saying — as best as I can figure it out — that the plan would close only around a third of the output gap over the next two years.

Here are the CBO's projections of the output gap (the difference between potential and actual GDP), assuming no stimulus:

(The graph is from these charts, which accompany the latest Budget and Economic Outlook)

Brad DeLong says:

I agree with Paul that this fiscal boost plan is too small, but I do want to admit that doing this well is harder than it looks. The tax-cut part does not look terribly effective as a stimulus--it is a step toward compensating for higher income inequality and a political play to make it more likely that Republicans will lose politically by trying to block the package rather than a significant boost to employment. Thus I do not think you would want to make the tax-cut part larger. And it is hard to find a lot of additional spending projects that can be ramped up quickly and do a lot of good--relatively soon in that endeavor the short-term fiscal multiplier falls below one. They are trying their best.

At Econbrowser, Menzie Chinn notes that, with the output gap expected to last for a while, we shouldn't be so concerned with the "lags" in implementing infrastructure spending (a point also made by Krugman in this column).

The implicit assumption here seems to be that the recovery will be slow, as it was following the 1990-91 and 2001 recessions. In both of those cases the unemployment rate continued to drift upward after the business cycle peak declared by the NBER. In terms of magnitude, it looks like we're now in a more serious recession, comparable to 1973-75 and 1981-82. Those recessions had bigger spikes in unemployment, but also much brisker recoveries. Unemployment peaked at 10.8% in December 1982 and fell to 8.3% at the end of 1983 - a decline of 2.5 percentage points in 1 year. The unemployment rate declined 1.6 points from May 1975 (9.0%) to May 1976 (7.4%). In both cases, the recoveries were quicker than what is envisioned by Romer and Bernstein or by the CBO.

It may indeed be that the 'jobless recovery' is the new 'normal' pattern. If the dynamics of earlier recessions were driven by manufacturing firms responding to unintended inventory accumulation, improvements in inventory management and the declining share of manufacturing might help explain the changing pattern. However, part of the reason economies can bounce back quickly (Milton Friedman once compared recessions to plucks of a guitar string) is that idle resources can be put back to work fast. And it looks like we will have no shortage of idle resources this time around.