Which Is The True Jobless Rate Correlation? Charting The Schrödinger Unemployment Rate
In an essay by Pimco's Tony Crescenzi, using the old and worn out title "To QE or Not to QE", which asks just that question, one of the lines of analysis focuses on the traditional conventional wisdom relationship between the jobless rate and initial claims for unemployment insurance. Tony says that this correlation leads him to believe that the unemployment rate is lower than where it official stands because, "Progress has been made, for example, on the employment front, with the six-month moving average for private payroll gains increasing to 214,000 per month in the six months ended in February 2012 from 160,000 per month in the 12 months prior. Importantly, weekly filings for initial jobless claims have fallen to a four-year low, fully 100k below year-ago levels and in territory consistent with a further decline in the unemployment rate (see Figure 1)."
So far so good, and indeed if one very simplistically tracks merely the unemployment rate to jobless claims, the picture does indeed seem rosier than it currently is. The problem however, is that as always happens in this case, initial claims reflect only a discrete component of the true unemployment situation in the New Normal, which more than anything is characterized by one specific feature: the avalanche like implosion of the labor force, and the departure of millions of people, almost monthly from the labor pool, noted so very often on these pages, and recently forcing even Goldman and JP Morgan to ask whether Okun's law is not in fact broken precisely because of this. As such there is one other correlation that in our humble opinion should be tracked far more closely when trying to anticipate the unemployment rate: that of the unemployment rate but not just to initial claims, but rather to initial and continuing claims, as well as extended benefits and EUCs, which provide a far better picture of those who are truly falling out of the labor pool. And as the chart below shows, when using that far more accurate New Normal correlation, the picture is decided worse. In fact, instead of a sub-7% implied unemployment rate, the true implied unemployment rate is just over 12.5%, precisely in line with our calculations if one assumes a historical labor force participation, instead of using election year gimmicks to paint the economic tape.
Which correlation is accurate: we leave it up to our readers to decide, but note that just like everything else in a centrally planned world, we now have yet another Schrödinger data point- what the propaganda machine wants it to be, and what it really is.