Inside the Numbers: Unemployment

This month's report has a some serious statistical anomalies. New jobs (36k) badly missed expectations (146k) as well as whisper (180k). At the same time, the unemployment rate inexplicably dropped sharply from 9.4% to 9.0%, the second steep drop in a row (it had been at 9.8%). This is the largest two month drop since 1958 (see chart, from Bespoke). How can this be, given the dismal job growth? 

As MIsh succinctly puts it, to get the rate down in two months from 9.8% to 9.0%, the BLS simply did not count 2M workers. The workforce needs to absorb about 1.5M new workers (eg college grads), but last year the BLS says the labor force fell by a net 167k, with 2.1M former workers simply dropping out. He concludes:

Look on the bright side, at this rate we will back to full employment in no time

It should be noted that the BLS uses two different surveys for these numbers. The new hires come from business surveys, and the rate comes from household surveys. While the business survey shows 36k new jobs, the household survey showed 117K new jobs, but 504k people either dropping out of the workforce or dropping out of the population! Yes, the survey said we lost 185k people out of our working-age population. This is assuredly wrong, a statistical anomaly coming from incorporating new census data. Without this change, the rate would have dropped to 9.3%. 

Also without this change, the new jobs would have been a much higher 589k, way out of line with other surveys. Gallup does their own survey, and it has shown an increase in the unemployment rate:

Weather was blamed for the mixed unemployment report. Outside jobs were down, like construction (-32k) and transport (-38k), while inside jobs like manufacturing were up (49k), its best rise in 12 years (!!), although non-durable manufacturing dropped (-13k). If weather is to blame, the February unemployment report should beat expectations as delayed hiring catches up. Look for 260k new jobs in Feb.

The simple explanation for the fewer-new-jobs but lower-unemployment-rate is the 99ers are disappearing from the stats faster than new jobs come in. 99ers are the unemployed who have maxed their benefits (the number refers to 99 weeks of benefits), and disappear from the official stats even if they are still seeking jobs.

We saw this last month as well, and it should continue for several months as the 99ers reflect the big job losses at the start of the crisis.

In the past year, the number dropping out has increased by 2.2M, while the number of "official" unemployed-seeking-work has dropped by 1.2M even though job growth remains below the 150K monthly rate just to absorb new graduates into the workforce. It should not surprise, then, that labor participation continues to fall, now at a 26 year low:

The fate of the 99ers is supposed to be captured in the U6 number, which includes discouraged job seekers. This month the seasonally-adjusted U6 fell a huge amount, from 16.7% to 16.1%. Good news, right? Nope, a drop like this is also a statistical anomaly. The non-seasonally adjusted U6 surged to 17.3%, and is now seriously out of whack with the adjusted numbers:

As you can see from the chart, something screwy happens at the turn of the year, and it will take several months to see if the seasonally-adjusted number is too low or the non-seasonally-adjusted number is too high. The seasonal adjustments include the impenetrable BLS birth/death model (of small business, not people), which during the Great Recession showed expansion of new businesses! They get the Darwin Award for that brilliance. The BLS did a major recounting of their birth/death model in January, and it now shows business destruction where they used to show creation.

When the BLS revises past numbers down, it makes the new numbers look better.  

Another statistical adjustment is last year's job creation numbers also get readjusted in January, and this time 400k jobs were removed from 2010! Given we added only 36k jobs in January, it was as if 12 x 36k were eradicated last year, removing a year's worth of Januaries. In total, 538k jobs have now been removed from the prior two years of stats. Hence the recoverlyless recovery was worse than it was reported. 

My take on all of this is we remain in a flat period and have been for a year. Cut through the noise and the household survey showed a 117k increase in January, much less than its 297k in December but very much in line with monthly numbers over the past year, and slightly less than we need to absorb new entrants to the workforce as the population grows.  

Cutting through the stats, take a look at the trend. The revised BLS surveys show a flat 2010 that has begun to trend down, not up, although I expect that to change in February due to the weather effect: 

You can also see this in the percent change chart, comparing this recession with others:

Sadly, we remain well above where Christina Romer and the Obama Administration predicted we would be after the Stimulus:

 

Comments

[Chabazite]

Good post, Yelnick. Just interested in anyone's prognosis on the following dilemma. $2T of QE (including the $600B since November) has had the effect of boosting share prices for the minority, and commodities which unfortunately impoverishes the majority. Meanwhile, said FED actions appear not to have turned the US money supply positive http://www.shadowstats.com/alternate_data/money-supply-charts so I guess it could be argued that the US remains stuck within a deflationary cycle. So a continuation with the policy, say with QE3, results in a devaluation of the currency until Joe Sixpack starts to scream 'no more' because he can't afford the basic necessities of life. How does this get resolved from ECONOMIC standpoint. Does it end up in some kind of deflationary collapse 'a la Prechter', or a rebasing and stabilising of the dollar around (say) gold? Interested to get views on how this gets resolved in the longer term.

[thoughts?]

I think that, at this stage in history, finance is like a top that is about to fall because it is losing angular momentum. We will wobble from side to side dramatically before falling to one particular side. Financially speaking, this would be more and more extreme and rapid inflations and deflations. If the stock market tumbles, for example, easy money will pour out even more freely than before in ever more desperate attempts to inflate... but the inflation will go directly and powerfully into the most essential things: energy and food. So we'll get, oh, doubling and tripling of oil and grains within days.

Dunno! Just seems to me that, if we are going into a deflation, it will be with dramatic, ridiculous inflations along the way. Kicking and screaming, so to speak. I mean, heck, the 30s is everyone's nightmare.

[Account Deleted]

AAPL Weekend Update: http://niftychartsandpatterns.blogspot.com/2011/02/apple-weekend-update.html

[Mamma Boom Boom]

>I mean, heck, the 30s is everyone's nightmare.<

How do you know? Where you there?

[?]

I have an idea.

Every single unemployed American receiving financial assistance in any way should be given a "job" to do nothing. Then since these are all new government workers (doing nothing) and now have jobs we can then say that the economy is back to full employment.

[Mamma Boom Boom]

? ....... It's not an easy puzzle, at this point. the American people have been sold down the river. Bill Clinton did it. I've been very vocal about that, made several post right here, if you remember.

It's almost hopeless, at this point. We don't have a congress that cares, so nothing will be done. The unemployed simply have to live with it, gather in communes, pool their resources, defend themselves from thieves and government thugs.

[Chabazite]

Every single unemployed American receiving financial assistance in any way should be given a "job" to do nothing. Then since these are all new government workers (doing nothing) and now have jobs we can then say that the economy is back to full employment.
---------------------
Yep - almost exactly what they do in India. Deliberate goverment policy. Give everybody a third of a job and pay them a third of a wage. That way no one starves.

[employment genius]

In last year's State of the Union, Obama declared job creation his "No. 1 focus," then spent much of 2010 on other priorities like overhauling healthcare and Wall Street rules.

With the elevated unemployment rate still ranking as Americans' top concern, there is little doubt jobs will again be the centerpiece of Obama's speech.

But more than ever before, Obama is also expected to use the annual address to cast himself as more of a fiscal hawk, possibly a tough sell for a leader presiding over trillion-dollar-plus annual budget deficits.

[yelnick]

chab, the Bernank's "wealth effect" for a few has created a "poverty effect" for the many - the squeeze in higher energy and food prices. I do not think QE3 is inevitable, especially if this stock market continues to climb into May when the decision on ending QE2 fast or slowly will be made. I expect in any event a pause between QE2 and QE3, which should make stocks stutter from June to Sept. If the commodities bubble does not crack by then the Fed may be forced to extend that pause.

[Mamma Boom Boom]

>especially if this stock market continues to climb into May<

Ah-ha, another bull!

[Chabazite]

Interesting Yelnick. Thank you for the response.