In lieu of an abstract, here is a brief excerpt of the content:

  • Comment and Discussion
  • George L. Perry

George L. Perry:

William Nordhaus's paper covers several related aspects of the current economic scene: the recent recession and current recovery, the behavior of profits during and leading up to the recession, and the prospects for the stock market in the recovery. He offers provocative ideas on all fronts.

Nordhaus compares the ten postwar recessions identified and dated by the NBER by examining four broad measures of economic loss: declines in real GDP, declines in employment, increases in unemployment, and positive output gaps as measured by the CBO. Output and employment declines track NBER-dated recessions very closely. Increases in unemployment lag these other measures of change, especially when measured over four quarters as in Nordhaus's figure 2, and this indicator yields several false-positive identifications of NBER recessions. The pattern of output gaps bears little relation to that of NBER recessions because these gaps usually remain positive for extended periods after a recession trough. However, if properly measured, the output gap would provide a summary indicator of overall economic loss.

From this broad-brush comparison of recessions, Nordhaus draws both a particular and a general message. The particular message is that last year's recession was the mildest of the postwar period and that at its trough, which he takes to be the fourth quarter of 2001, the economy was virtually at its potential output. I will say more about that later. The general message is that the NBER's approach to recession dating is outdated: recessions differ so much in their severity and length that a richer classification scheme is called for; meanwhile advances in economic science and [End Page 221] data availability allow us to do a better job of classifying recessions than the pioneers of business cycle measurement were able to do. Nordhaus suggests developing quantitative criteria that would allow business cycle downturns to be classified more like hurricanes are, with several gradations reflecting their severity.

Nordhaus gives the impression that a lot is at stake in refining our definition of recessions and applying more data to the task of identifying and classifying them. He hints that business-cycle dynamics could reflect the propagation of small impulses into large outcomes, so that identifying incipient virulent recessions would have important benefits. As a scientific matter, that is an important question and related to the more general question of whether periods of recession are special, in the sense that agents respond differently in recessions than at other times. That would be a reason to care a lot whether we were in or heading for such a period. If we could predict emerging weakness better, we should. But that goal is and always has been hotly pursued. It has nothing to do with classifying downturns according to severity after the fact.

Turning to that more modest task, I agree with Nordhaus that the NBER should look not only at changes in but also at levels of economic activity in identifying recessions. And that criticism holds true even if we do not expand the NBER dating committee's limited mandate, which is to identify "a significant decline in activity spread across the economy " (emphasis added). I have always believed that one strength of recession dating by a committee of experts was that they were free to consider any and all evidence of recession. There were surely stretches in the first half of the 1930s when employment rose, but we all speak of that whole period as the Great Depression. And if we took away several months of dead-cat bounce in the economy that followed its precipitous decline in the fi rst half of 1980, the whole of 1980-82 would be one recession, not two as judged by the NBER. It felt like one recession at the time. I see no reason other than precedent for the experts to look only at whether measures such as sales and employment are falling in identifying when a recession starts and, more important, when it ends.

However, going from principles to practice, I find the CBO output gap that Nordhaus uses in analyzing the level of economic utilization to be inadequate. The main problem...

pdf

Share