Reading this morning’s latest weekly update on jobless claims inspires the question: When will we see evidence that a new recession is here, or lurking in the near future? The answer: Not today. If there’s a clear sign that the economy’s set to tumble, it’s not obvious in last week’s new applications for unemployment benefits. In fact, this leading indicator continues to tell us that the labor market is slowly improving. New claims dropped by 12,000 to a seasonally adjusted 367,000 last week. One number doesn’t tell us much, of course, but it’s hard to dismiss the trend.
As the chart below shows, new jobless claims have been zig-zagging lower for months. As of last week, claims are near a four-year low. You can’t rely on any one indicator for business cycle analysis, but jobless claims have a fairly good record over the last four decades of providing an early warning sign of recessions. The fact that the new-claims trend shows no sign of rising can’t be accepted as the last word on what comes next, but this indicator surely increases the pressure on economists who argue that there’s a recession in our midst.
Some analysts say flat out that we should dismiss initial claims because of seasonal factors. But that view isn’t convincing. The year-over-year change in unadjusted claims supports the seasonally adjusted figures. Indeed, as the second chart below reminds, new claims have been routinely falling by around 10% a year since last spring before adding a seasonal adjustment.
Perhaps the debate should focus on whether jobless claims are fundamentally delivering the mother of all misleading signals for reasons that go behind technical issues. Has this indicator’s relationship with the business cycle changed? There are precious few absolutes in macro and so we need to keep an open mind. As the saying goes, this is economics, not physics. But the argument that jobless claims are no longer relevant at this point would be stronger if they were a lone signal of optimism. But as we learned yesterday, manufacturing seems to be improving too in the new year. A big part of the improvement is related to higher auto sales. As Bloomberg reports: “Automakers sold new cars and trucks in January at the fastest pace since the 2009 ‘cash for clunkers’ program without resorting to profit-sapping discounts, signaling demand returned to pre-recession levels.”
True, job growth is still sluggish, as yesterday’s ADP Employment Report advises. But the labor market has been sluggish for several years and the economy has still managed to expand, albeit modestly and in fits and starts. Has something changed? Maybe, but you won’t find a smoking gun for that view in today’s jobless claims numbers.
“It certainly suggests we will continue to see job growth at the higher end of the recent range (which has been between) 100,000 to 200,000,” says Christopher Low, an economist at FTN Financial. “If claims continue to drop then we should see job growth stronger than that.”
But if the claims data is wrong this time, and there’s a recession approaching, as a handful of economists warn, there will be clear signs of a downturn in the labor market. Employment trends aren’t considered a leading indicator, but at some point there’ll be a clear break from the recent trend of modest job growth to something more ominous. Perhaps we’ll find some clarity in tomorrow’s payrolls report for January. For what it’s worth, the crowd’s expecting more of a gray area rather a decisive number one way or another. The consensus forecast anticipates a net rise of 168,000 for private nonfarm workers, according to Briefing.com. That’s not high enough to put the recession fears to rest, although it’s not low enough to give the recession forecasters a victory either. If 168k turns out to be accurate, what it will do is keep the debate alive about the next phase for the business cycle.
“This is a mild positive, but with the market at these lofty levels, you need to have continued good news for the market to sustain its gains,” says Beth Ann Bovino, senior economist at S&P. “This is not a particularly meaningful data point, but it is more good than bad.”