April 9, 2012
A Bit Of Humility Is Still Required For Predicting The Business Cycle
Was Friday's disappointing news on job growth the death knell for the recent run of economic growth? Some analysts think that the macro jig is up and that the sharp slowdown in the expansion of payrolls for March represents the point of no return. The truth is that no one knows for sure if last month's downshift was a sign of things to come or just a temporary weak patch in an otherwise reviving labor market. In contrast to the far more widely followed establishment profile, the so-called household survey of the labor market in March certainly paints a more encouraging picture, as Scott Grannis explains. In any case, for now it's a mystery that can only be solved with more data. That doesn't mean we should ignore the potential for trouble--the latest establishment survey numbers are certainly disturbing at a time when growth overall remains fragile. The risk for a slowdown or worse suddenly looks higher. But it's premature to argue with a high degree of confidence that a new recession is unavoidable.
Consider, for instance, how the annual growth rate for private-sector nonfarm payrolls compares with initial jobless claims. Even after the weak jobs report for March, the year-over-year increase for payrolls is a respectable 1.93%, or near the highest levels reported since the last recession was officially declared over in mid-2009. If there's a new recession knocking on our door, payrolls growth will deteriorate rapidly and dramatically in the months ahead.
That may be coming, but it's not obvious from looking at the trend in initial jobless claims. One of the stylized facts about recessions is that new filings for unemployment benefits rise persistently just ahead of, or in the early stages of recessions. But there's no sign of that warning--not even close. Initial claims have been falling fairly consistently for nearly a year. The trend, if history's a guide, signals that the labor market will continue growing.
It could all fall apart suddenly, of course. But while we're waiting for fresh data, there are several economic data points to consider that don't make it easy to forecast a new recession. For instance, average weekly hours worked among production and nonsupervisory employees has been rising steadily. Looking beyond the labor market, industrial production's year-over-year growth rate looks robust and stable. A broader measure of U.S. economic activity shows no clear signs of trouble either, as per the Chicago Fed National Activity Index. Meantime, one measure of financial stress continues to recede, based on the St. Louis Financial Stress Index. And in contrast to March's slowdown in employment growth, manufacturing activity perked up a bit last month. The services sector continued growing last month too, albeit at a moderately lesser pace.
Still, this is no time for complacency. One warning sign that's been a regular on these pages for months is the decelerating growth in disposable personal income. Other risk factors include high energy prices and the ongoing euro crisis.
Overall, you can make a good case for worrying about the economy, but that's been true all along. But it's too early for a high-confidence forecast that a recession is upon us. Predicting major turning points in the business cycle remains a speculative affair if the expected change isn't conspicuous in the data. For all the progress in macroeconomics in reading and analyzing big-picture trends, the best you can expect is that we'll have a defacto confirmation of the onset of the next recession in the early stages of that event—not before it's started.
For instance, in late-March 2008 I went on record in arguing that a new recession was unavoidable. There were several factors that convinced me that the cycle had irrevocably turned for the worse, although lots of analysts disagreed at the time. The official confirmation arrived months later, in December 2008, when NBER announced that a new recession began roiling the economy as of January 2008 (the cyclical peak was December 2007). By that standard, my March 2008 recession call was late, which is to say well after the recession had already begun. A number of economists were warning of a downturn even earlier. But in March 2008, and for several months thereafter, anticipating a recession was still a minority view.
The point is that at some point it becomes a virtual certainty that another contraction is inevitable and so we should focus on making that call as early as possible. The numbers will tell us when that point of no return has arrived. For the moment, though, we're not there yet. We may be in a recession, but that's still a speculative call without sufficient statistical confirmation.
On that note, the week ahead won't change much. Thursday brings another weekly update on jobless claims, but one number for this volatile series doesn't usually tell us much. The March report on consumer inflation arrives on Friday, but that's of minimal value for assessing the broad economic trend.
In short, it's going to take time—another month or two at a minimum—to decide if the economy's set to tumble. Meanwhile, let's keep in mind what's obvious. The economy wasn't in recession in February, and probably didn't succumb in March, based on the preliminary data in hand so far. Confidence levels fall sharply for making claims about the cycle for April and beyond. It's still impossible to see around corners no matter how hard you look in the rear-view mirror. We should be looking, of course, but it's important to recognize what mere mortals are capable of in the dark art of forecasting the business cycle.
Posted by jp at April 9, 2012 8:09 AM