It takes spine to sell an economic recovery on a net jobs gain of 46,000 since December 2008. So hats off to the president and his handlers. They’re gutsy.
My firm uses software and mathematics to track equity-market behaviors. We’re statisticians, recognized experts regularly quoted on market matters in major media. We like data.
On November 2, the Bureau of Labor Statistics (BLS) reported that unemployment crept up to 7.9% but the economy created more jobs than expected. In equity-trading markets, you can watch tick-by-tick moves or you can apply broader behavioral brush strokes like we do. The “labor tick” looks good. But not the big behavioral measures.
In December 2008, the BLS reported (visit the BLS archives here to see):
Adult population: 235.03 million
Adult workforce: 154.45 million
Employed Adults: 143.34 million
Unemployment rate: 7.2%
In October 2012, the BLS reported:
Adult population: 243.98 million
Adult workforce: 155.64 million
Employed Adults: 143.38 million
Unemployment rate: 7.9%
Look at the number of employed adults. It’s barely — 46,000 — higher now than in December 2008. The president claims 5.5 million new jobs. The BLS household survey from dark nadir to weak zenith in the presidential term may show gross additions near that number.
But the president doesn’t tell everybody that the economy simultaneously lost the same number of jobs. He doesn’t mention that the adult population has increased by 9 million (subtract the adult workforce above in October 2012 from that in December 2008) or that the workforce has only expanded by 1.2 million over the same time.
These aren’t my figures. They’re the BLS’s. The data in simplest terms say that just one of every eight or nine adults capable of joining the workforce has done so since December 2008. Where are the rest? The explosion in student loan balances and social safety net expenditures implies that they’re going to school or living on welfare.
Here’s another stubborn and inconvenient fact. If we use the statistical measures applied in the second term of President George W. Bush, the unemployment rate is 11.2%.
It’s not that it should be 11.2%. It is 11.2% if we apply identical mathematics to each data set. The reason the unemployment is 7.9% instead of 11.2% is because different metrics have been used.
Let me explain. Going back to the 1990s in the BLS data, the ratio between adult population expansion and workforce growth is consistent, modified by demographic trends. For instance, moving into the 2000s, the effects of retiring baby boomers halted expansion (Image 1) in workforce participation tracing to roughly 1972 (after the U.S. abandoned the gold standard and the dollar began to depreciate rapidly, and more had to work to sustain living standards).
But step to January 2009, and we find the root of a Butterfly Effect — minute change fostering consequential outcomes. To illustrate (Image 2), we totaled the adult-population gains from January 2005 to December 2008, the second term of George W. Bush. The adult population increased 10.2 million. The adult workforce rose 6.5 million. That’s a ratio of 63.4% — numerically akin to reported “participation rates” but not the same. It means two-thirds of the new adult population joined the workforce.
Compare that to January 2009-October 2012. The adult population increased 9.2 million while the workforce was up only 1.9 million, an astonishingly insipid ratio of 20.8%. The rate is even worse if it’s traced to December 2008.
Generally, the prevailing ratio of population to workforce has varied little over the decades. The average variance between the two in 2008 was 0.037%. But full-year 2009 variance jumped to 0.137%. That seems small but it’s monumental. By definition, a change of more than three standard deviations is a Black Swan — a Leviathan in your data.
What happened in 2009? The stimulus. What about the economic collapse? That happened in latter 2008 — no similar statistical Black Swan there. The data suggest that rather than saving jobs, the stimulus shrank the work force. There goes the “it could have been worse” argument.
The ratios in 2011 and 2012 then widened by a remarkable 42% over standard deviation, pointing to falling participation relative to population growth. Historically, participation bounce-back is a hallmark of recoveries. It was true circa 1964, 1972, 1980, 1992, and 2004 (see Image 1). In each instance, declines or plateaus beforehand in workforce-participation reflected economic struggle. Participation then expanded with recovery.
But participation remains down. What’s more, even with better recent reports the trend line of monthly employment data from January 2009 to October 2012 is -0.28%. Statistically over this president’s term, we continue to lose jobs faster than we’re creating them.
That’s what the math shows. What happens if we continue the same policies that got us into this mess?
Tim Quast is founder and president of ModernNetworks IR LLC in Denver.