Since writing about certain unemployment data that has the uncanny ability to signal the end of recessions, I found this post from the economist who discovered the phenomenon: Robert Gordon of Northwestern University.
Although dated May 1, prior to my original post, it has only surfaced in critiques in recent days. And it deserves to be aired, because in it Mr. Gordon puts forward his startling forecast:
"It is always too early to make definitive conclusions, but the recent 2009 peak in new claims looks sufficiently similar to previous recession peaks to allow a conclusion that it is highly probable that the new claims peak has now occurred.... My reasoning leads me to conclude that the ultimate NBER trough of the current business cycle is likely to occur in May or June 2009, substantially earlier than is currently predicted by many professional forecasters."
Economist goes wild
Don't let the bland language fool you. That's the economist's equivalent of an underdog quarterback, say, Joe Namath of yesteryear, predicting victory in the Super Bowl. (You can read Gordon's full post here.)
Gordon has the creds to make such a prediction. Since 1978 he's sat on the Dating Committee of NBER, the National Bureau of Economic Research. (I know, it's a sexy-sounding title but the only thing this committee seems to date are the economy's peaks and troughs.)
His reasoning goes like this: The four-week average of initial claims for unemployment insurance reaches a peak in recessions about a month before the recession actually ends. The numbers have held up in all five of the previous recessions. The current four-week average peaked April 4, and has since declined for four straight weeks.
An exception from the '70s
While that's a remarkable correlation, I pointed out in my original post (which you can read here), that it wasn't foolproof. The average peaked five months too early in the 1969/70 recession. In his post, Gordon discounts that recession:
"[T]he second peak was an artificial outcome of a special event, the famous General Motors strike that lasted 67 days and began on 14 September 1970," he writes. "The initial peak of new claims suggests that the recession would have reached its trough at some point in June or July, but the direct and indirect impacts of the strike delayed the actual NBER trough until November 1970."
That makes sense statistically. But recessions ebb and flow, often because of such big and unpredictable events. If bankrupt Chrysler has to liquidate this year, that could well delay recovery. That's why recessions' finales are so hard to predict.
A peak too low?
The other criticism that other bloggers and I have pointed out is that, when measured as a share of the workforce, the current April 4 peak in jobless claims is quite low. For example, it's not quite as high as the record set in 1982, even though the workforce is 40 percent bigger today.
That makes many bloggers, like Tim Iacono, skeptical that the worst is over: "For example, to reach the 674,000 October 1982 peak for new unemployment insurance claims, we'd have to see a figure of over a million today. To equal the February 1975 peak of 561,000 would require over 1.1 million. That's almost double the recent peak!" he wrote in a recent post.
Gordon acknowledges that the numbers sound too low for a recession as serious as the one we're having: "[W]e are cautioned by the [data] showing that new claims in 2009 are running well below both 1973-75 and 1981-82 as a percentage of total insured employment."
But he's sticking to his guns. If he's right, that would be surprisingly bullish news for our globally battered economy.