An Old Forecast Gets A New Close-up… Again

On Tuesday, Lakshman Achuthan, the co-founder of the Economic Cycle Research Institute, made another visit to Bloomberg TV and insisted that last year marked the beginning of a new recession for the US. He once again found himself in the unenviable position of arguing that the numbers published to date don’t accurately reflect reality. Data revisions, he asserts, will eventually come around to his view of the business cycle and set the record straight..

In theory, he may yet find vindication, although that assumes a rather large and widespread batch of downward revisions. To be fair, Achuthan’s long-running forecast that the US economy will suffer sluggish growth for an extended period has been accurate. Of course, many analysts have been saying no less for several years. Granted, there have been times in recent history when an indicator or two has popped higher, inspiring hope that a stronger pace of growth had finally arrived. But so far, those enticing instances have been the exception to the modest growth that has continued to prevail.
Yesterday’s better-than-expected GDP report for the second quarter doesn’t change much. The crowd may have been pleasantly surprised, but no one will confuse recent economic activity as robust. Meanwhile, the Fed just downgraded its outlook for the US economy to “modest” growth. But while there’s much to lament about the pace of growth, there is growth, even if it’s lacklaster and unsatisfying for millions of Americans looking for work. In other words, there is a distinction between a slow expansion that leaves the economy vulnerable to all sorts of macro troubles and a recession that’s worthy of the name.
As for ECRI’s new new recession forecast, which will be two years old come this September, the numbers still don’t provide much evidence that a recession has already started, or that we’re in imminent danger of a downturn. As economist Tim Duy writes of Achuthan’s latest pronouncements on Bloomberg TV, he’s “desperately clinging to his call that a recession started in the middle of 2012.” Why “desperately”? Because the data tells us a different story. In particular, four primary data sets that have been front and center in NBER’s decisions in the past for calling cyclical peaks and troughs continue to trend higher:

A similar analysis is easily found in a broader set of indicators, as this recent update to The Capital Spectator’s Economic Profile reminds. What are we to make of ECRI’s ongoing campaign to convince a skeptical world that the US has been in a formal, NBER-defined recession since 2012?
Perhaps the only thing that you can truly count on in the delicate art/science of anticipating business cycle risk: We’re all destined to be wrong at times. That’s the nature of the beast. No one really understands the full scope of cause and effect for virtually infinite connections that comprise the $17 trillion US economy. If it were otherwise, we wouldn’t be having this discussion.
Even the relatively easy task of deciding what the published numbers tells us about the state of macro in the recent past can be a hornet’s nest of complication and misleading signals. It only gets worse when you try to see around corners. Pity the folks who like go far out on this already weak limb and routinely dispense predictions about what will be unfolding a year or two down the road.
ECRI’s problem, as I see it, isn’t that they’re wrong. Indeed, those of you with perfect macro forecasts and analyses please raise your hands. No one? Well, of course not. Rather, the issue here is insisting, for the better part of two years, that a forecast that sounded reasonable at first ultimately went soft. Heck, it happens to everyone, and that’s not going to change anytime soon.
The good news is that there’s quite a lot to be learned from mistakes, assuming, of course, one can admit that such things are possible. This sensible grasp of reality is particularly useful in the world of econometric-based forecasting, where analyzing the forecast errors can sometimes help improve prediction accuracy going forward. In the ongoing quest to build better models, integrating past mistakes into the analysis may actually be helpful.
More generally, learning from our mistakes is at the heart of progress, slow as it is, in macroeconomics. What we thought was true yesterday may not hold up to the light of investigation tomorrow. In fact, you can count on it in some degree. If there’s any hope of squeezing something productive out of mistakes, it begins by admitting them when they strike, and then figuring out what went wrong as a tool for improving the model.
Imagine if ECRI embraced what has become clear since September 2011: the economy has continued to grow, weakly, but it still looks like an expansion. It would surely be helpful to hear what ECRI thinks in terms of why it’s forecast didn’t quite hold up. Why did the economy continue to skirt recession, albeit at relatively low and vulnerable rates of growth?
Perhaps the framework to consider is akin to what happens when an airliner crashes. The National Transportation Safety Board sends out a team of investigators to figure out the hows and whys that created the tragedy. Increasingly, the smoking gun(s) are linked to human error, which probably isn’t too far from the truth in macro mistakes.
Some folks say that we would have had a recession if it wasn’t for the Fed’s extraordinary monetary interventions over the last two years. By this reasoning, we’re told, it’s unfair to say that ECRI has been wrong. But even if you think that the Fed kept us out of a recession (where have we heard that before?), it’s the job of economists to adjust their models and forecasts according to available and every-change mix of evidence. For most of us, the only goal is making relatively timely, reliable calls on peaks and troughs. Ascribing blame and looking for jokers in the deck is another matter entirely, and one that’s not all that relevant if you’re managing money and making real time decisions about your career, your business, etc.
Despite all this, let’s recognize that ECRI remains a serious consultancy that produces high-quality research, in no small part due to the impressive pedigree that laid the firm’s intellectual foundation, namely, the work of Geoffrey Moore. This economist, “a protege of Wesley Clair Mitchell and Arthur F. Burns, the founders of formal business cycle analysis, added much pioneering work of his own,” as The New York Times wrote of Moore after his death in 2000. Previously, “Before there was a committee to determine U.S. business cycle dates, Moore decided all those dates on the NBER’s behalf from 1949 to 1978,” according to ECRI’s web site. Moore, of course, brought his intellectual capital to ECRI, which was launched in 1996. The firm boasts that it’s refined Moore’s pioneering work on dating the business cycle in its ongoing efforts at identifying peaks and troughs. Nonetheless, it’s intriguing to wonder what Moore would say about recession risk these days.
Speaking of cycle dates, formal confirmations from NBER on the start of new recessions, Achuthan says in his latest Bloomberg TV interview, can arrive as long as two years after a downturn has commenced, or so NBER’s record on announcing cycle dates suggests. By that reasoning, and using ECRI’s claim of mid-2012 as the alleged start date for a new recession, we can look forward to another year of media interviews aimed at convincing the crowd that things really are worse than they appear.