Yves here. In case you missed it, there was a great deal of consternation at the surprisingly low level of job additions in Friday’s employment reports. Bloomberg put consensus at 189,000, when the report came in at 87,000 . And pretty much every other metric was not good either: the average workweek fell, the increase in average hourly earnings was at the low end of the range of forecasts. Oh, except the headline unemployment rate, which fell to 6.7% due to a jump in people leaving the official workforce. The weather is being blamed, but how does that wash when Sandy didn’t produce any blip in employment figures?
Some useful takes, in addition to Hugh’s usual thorough analysis:
The December Jobs Report: Disappointing But Not Surprising Robert Prasch, New Economic Perspectives
What Caused the Crash In the Labor Participation Rate? George Washington
El-Erian: Jobs Data Between Puzzling and Worrisome Bloomberg. And if you can stomach it, a later video of administration spokescritter Jason Furman with El-Erian, selling the party line that there’s noise in the data and you need to keep your eye on the trend. But OMG how can anyone buy a used car from this guy?
By Hugh, who is a long-time commenter at Naked Capitalism. Originally published at Corrente. A complete archive of Hugh’s reports can be found here.
The short story is this. In the Household survey, seasonally adjusted unemployment fell an incredible three-tenths of a percent to 6.7% in December. The labor force is around 155 million. 0.3% of that is around 450,000 workers. In fact, it was 490,000. Yet the Business survey reported, also seasonally adjusted, an increase of just 74,000 as compared to the upwardly revised November number of 241,000.
How to explain this wild divergence in the seasonally adjusted data? The number of employed grew by 143,000 (still twice the number of jobs reported created) while the labor force shrank 347,000. Add these two number together and you get 490,000.
It’s important to remember that the BLS uses a jobseeker definition for unemployment. If you are without a job and have not looked for one in the 4 weeks before the week in which the Household survey is conducted, you are not considered unemployed. You are considered to be out of the labor force. As a worker, you cease to exist. What happened in December is that a lot of workers ceased to exist.
The unadjusted numbers tell a similar story, only more extreme. The seasonally unadjusted labor force declined by 638,000, with employment falling 352,000 and unemployment, 287,000. The decline in employment corresponds to a 246,000 decrease in the number of jobs seasonally unadjusted from the Business survey.
This is a piss poor jobs report. The drop in unemployment is completely misleading. While December is not usually a stellar month, these numbers are worse than usual. Expectations in the markets were that this was going to be a gangbuster report. It was instead just a bust. This is a bad report.
Two notes: First, in the reports for December and January, there are various revisions to the surveys that confuse matters even more than is normally the case. Second, next month in the report covering January actual (unadjusted) employment and jobs will plummet following the holiday period by 1.5-2 million. However, the official (trendline) numbers bridge this abyss and often show healthy increases which do not, in fact, occur until later in the spring.
_________________________________________________________
Household/Employment Survey
Potential Labor Force
In December, the nation’s potential labor force, the Civilian Non-Institutional Population over 16 increased 178,000 from 246.567 million to 246.745 million. Multiplying this by the employment-population ratio (58.6%) yields 104,000, a rough indicator of the number of jobs needed to keep up with population growth. The 74,000 jobs number falls below this figure for the first time in quite a while.
It is important to remember in assessing job creation during the Obama Administration that in absolute terms there are still 1.390 million fewer jobs now than at the peak in November 2007, just before the recession began. And that peak was itself the culmination of six years of poor job growth during the Bush Administration. Now add in an increase in the potential labor force since November 2007 of 13.806 million and apply the current 58.6% employment-population ratio to it (8.09 million) or 62.9%, the one from November 2007, (8.684 million), and the estimate of the Obama jobs deficit from the November 2007 peak, again nothing to write home about, is, six years on, in the neighborhood of 9.5-10 million. None of this touches on the overall poor quality of the 7.517 million jobs created under the Obama Administration since the seasonally adjusted bottom in February 2010. But it does show that at the beginning of his sixth year in office in terms of the raw numbers, Obama is less than half way back to pre-recession levels. Add in the issue of quality, and you begin to see how abysmal Obama’s jobs record is.
Labor Force
Seasonally adjusted, the labor force declined 347,000 from 155.284 million to 154.937 million.
Unadjusted, it decreased 638,000 from 155.046 million to 154.408 million.
Participation Rate
Seasonally adjusted, the participation rate, the ratio of the labor force to the potential labor force, fell two-tenths of a percent to 62.8%, the same as in October.
Unadjusted, it fell three-tenths of a percent to 62.6%.
Employment
Seasonally adjusted, employment grew 143,000 from 144.443 million to 144.586 million.
Unadjusted, it fell 352,000 from 144.775 million to 144.423 million.
Employment-Population Ratio
Seasonally adjusted, this ratio was unchanged at 58.6%.
Unadjusted, it declined two-tenths of a percent to 58.5%. For the year, it was 58.6%, the same as last year.
Unemployment
Seasonally adjusted, unemployment fell 490,000 from 10.841 million to 10.351 million.
Unadjusted, it dropped 287,000 from 10.271 million to 9.984 million.
Unemployment rate
Seasonally adjusted, it declined three-tenths of a percent to 6.7%.
Unadjusted, it fell one tenth of a percent to 6.5%.
Full Time vs Part Time Employment
Seasonally adjusted, full time employment increased 327,000 from 116.951 million to 117.278 million. Part time employment decreased 89,000 from 27.461 million to 27.372 million.
Unadjusted, full time employment fell 214,000 from 116.875 million to 116.661 million. And part time employment decreased 138,000 from 27.900 million to 27.762 million.
Involuntary vs. Voluntary Part Time Employment
Seasonally adjusted, involuntary part time workers were little changed increasing 49,000 from 7.723 million to 7.771 million. Voluntary part timers decreased 127,000 from 18.858 million to 18.731 million.
Unadjusted, involuntary part time employment grew by 427,000 from 7.563 million to 7.990 million. Voluntary part time workers decreased 434,000 from 19.628 million to 19.194 million.
_________________________________________________________
The U-6
The departure of workers from the labor force was not captured by the seasonally adjusted U-6, the BLS’ broader measure of un- and under employment, which remained unchanged at 13.1%.
The seasonally adjusted U-6 was composed of 10.351 million unemployed, 7.771 million involuntary part time workers, and 2.427 million of the marginally attached (those who have no job but looked for work in the last year but not the last month; an increase of 421,000 from November), or 20.549 million, a decline of 173,000 from last month. Basically, the decrease in unemployment was more or less cancelled out by the increase in the marginally attached.
The unadjusted U-6 increased three-tenths of a percent to 13%.
[Standard note]
The BLS uses a restrictive job seeker definition of unemployment, that is without a job but have looked for one in the last 4 weeks. The marginally attached are not counted as part of the labor force and their use in the U-6 is an indication that this is what the BLS considers its functional undercount to be.
The BLS also has a more extended category: Not in Labor Force, Want a Job Now (seasonally unadjusted). In December, this increased 495,000 to 5.932 million. So this measure does capture some although not all of the unadjusted 638,000 decline in the labor force.
This BLS category does not often reflect well actual movements in the economy. So I have developed a simple alternative to it. I calculate the size of where the labor force should be by multiplying the potential labor force of the NIP by a participation rate characteristic of a solid economic expansion (67%, the Clinton boom was at or above this level for nearly 40 months). The difference between this and the current labor force measures the size of the real BLS undercount, those who do not have jobs but would work if jobs were available to them. This then allows me to recalculate where real unemployment is and where real un- and under employment (disemployment) is.
Next month for the 2014 data, I will begin revising, probably on a yearly basis, this 67% figure in light of the effects of Baby Boomer retirements. I should note that there are offsets to these and I expect revisions to be in the range of 0.1-0.3% a year.
.67(246.745 million) = 165.319 million (where the labor force should be)
Trend Undercount:
165.319 million — 154.937 million = 10.382 million , an increase of 476,000 from November
Current Undercount:
165.319 million — 154.408 million = 10.911 million, an increase of 757,000
_________________________________________________________
Real Trend Unemployment (that is seasonally adjusted) :
10.351 million (U-3 unemployment) + 10.382 million (undercount) = 20.733 million, down 80,000
20.733 million / 165.319 million = 12.5%, down 0.1%
Real Unemployment Now (i.e. seasonally unadjusted) :
9.984 million (U-3 unemployment) + 10.911 million (undercount) = 20.895 million, up 470,000
20.895 million / 165.319 million = 12.6%, up 0.2%
_________________________________________________________
Real Trend Disemployment:
Real Trend Unemployment + involuntary part time workers seasonally adjusted = 20.733 million + 7.771 million = 28.504 million, down 28,000
28.504 million / 165.319 million = 17.2%, down 0.1%
Real Disemployment Now:
Real Unemployment Now + involuntary part time workers seasonally unadjusted = 20.895 million + 7.990 million = 28.885 million, up 897,000
28.885 million / 165.319 million = 17.5%, up 0.6%
Analysis
What we are seeing here is the trend exhibiting some improvement while the actual numbers are showing the beginning of the post-holiday dropoff which will be accentuated further next month.
Other
The number of long term unemployed (6 months or more), as defined within the BLS job seeker model, decreased 166,000 to 3.878 million. The long term unemployed account for 37% of the U-3 unemployed, unchanged from last month.
White unemployment decreased two-tenths percent to 5.9%. White teen unemployment increased 0.7% to 18.0%. African American unemployment decreased 0.5% to 11.9%. African American teen unemployment decreased 0.2% to 35.5%. As always, the rule of thumb is African American unemployment is twice that of whites.
_________________________________________________________
Establishment/Business/Jobs Survey
Jobs
Seasonally adjusted, the number of private sector jobs increased 87,000 and decreased 13,000 in government netting the reported weak 74,000 increase. November jobs were revised upward 38,000 to 241,000. October was unchanged.
Seasonally adjusted, total nonfarm jobs increased 74,000 to 136.877 million. Total private jobs increased from 114.941 million to 115.028 million.
Unadjusted, total nonfarm jobs fell 246,000 from 137.999 million to 137.753 million. Total private jobs decreased 120,000 from 115.662 million to 115.542 million. Government jobs declined 126,000 from 22.337 million to 22.211 million.
Unadjusted, where the economy is now, construction lost 216,000 jobs to 5.745 million, accounting for most of this month’s job losses. Manufacturing was largely unchanged, gaining 4,000 to 12.028 million.
Unadjusted, the super sector private service-providing increased 92,000 to 96.883 million, of which wholesale gained 17,100 to 5.8211 million, retail grew 176,500 to 15.9452 million, information dropped 24,000 to 2.673 million (motion picture and recording lost 30,900 to 347,500), professional and business services lost 44,000 to 18.858 million (while temp services gained 17,500 to 2.8893 million. Healthcare added just 1,800 to 14.6997 million. Leisure and hospitality lost 53,000 to 13.963 million.
Hours and Earnings
Average weekly hours for all employees on private nonfarm payrolls fell one-tenth hour to 34.4 hours (a tenth hour less than a year ago). Average hourly earnings increased 2 cents to $24.17/hour, but average weekly earnings fell $1.73 cents to $831.45. for a 2.1% increase yoy.
Average weekly hours for production and nonsupervisory (blue collar and clerical) personnel also fell 0.1 hour to 33.6 hours (also a tenth hour less than a year ago). Average hourly earnings increased 3 cents to $20.35/hour. Average weekly wages fell $1.02 to $683.76, a 1.5% increase yoy.
Household data (Employment/unemployment)
Statistical significance: +/ – 300,000
The A tables: http://www.bls.gov/cps/cpsatabs.htm
A 1 for most information and categories
A 2 Unemployment by race
A 8 Part time workers
A 9 Full time workers
A 12 Duration of unemployment
A 15 U 6 un- and under employment
A 16 Persons not in labor force
Establishment date (jobs)
Statistical significance: +/ – 90,000
The B tables: http://www.bls.gov/ces/cesbtabs.htm
B 1 Total jobs and jobs by industry/type
B 2 Weekly hours, all employees
B 3 Hourly and weekly earnings, all employees
B 6 Weekly hours, blue collar
B 7 Hourly and weekly earnings, blue collar
19 comments
Comments are closed.
I greatly admire Hugh for addressing these figures month after month after month, but somehow I think all this talk about an ‘unemployment rate’ obscurantist. In a country of 300 million people, I suspect there are no more than10-15 million who can continue to exist without the support of some kind of job. If we posit an average family size of 2.5, itself something of a stretch, the country would need 114 million jobs capable of providing for 2.5 people each, and the private sector barely achieves that many jobs in toto, and what percentage of those are inadequate to support even one person without supplementary assistance?
A lot of supplementary assistance is provided through family and friends.
For those still able to do it, some assistance comes through slowly inching up on their credit card debt, which with regular payments credit card companies will up the debt limit and impose 20% compounded rate ad infinitum.
There are sell-offs of household goods through yard sales. Businesses conducted through cash payments on a casual basis. Exchanges of favors and gifts.
All those things that conservatives see as the great savior of the unemplyed in fact do exist but far from the levels of support that are necessary. And the philanthropic charitable giving of private NGOs and foundations is a big huge tax-deductible joke that lets the well-off write off their socialized art and music and culture purchases and feel good about “experimental pilot programs” that never can scale up.
There are thousands of foundations and non-profits outside of art, music and culture. A lot of non-profits tackle hunger, poverty, medical care, etc.
There are a number of unusual issues around employment this November and December. First of all, unemployment insurance was not extended by Congress. This has an effect two ways. (1) People receiving benefits must show they are actively seeking work in most states by demonstrating that they have done certain tasks during the week. Why would those who had lost their benefits bother to be actively seeking work with the same intensity? (2) November through January are typically horrible times for finding work because those who are hiring are on mental if not actual vacation. People who have been through the unemployment drill several times, and who hasn’t these days, tend to take those months off to try to restore some optimism.
In addition as the jobless “recovery” drags on it gets much harder for people to maintain the optimism required to keep being told that there are no jobs.
And the continued austerity (or is it economic sabotage) of Congress is familiar at the grassroots. For Republicans it will continue because they think that a bad economy helps them retake the Senate; it worked in the House. For the Democrats, dodging the “tax and spend” label seems more important than anything else and they are quite willing to lose in order to dodge responsibility.
The BLS is merely the messenger. The spin doctors who are trying to put a happy face on this sucky economy could save us money on the deficit by resigning. Sweeping layoffs of government PR flacks is several decades overdue.
“Why would those who had lost their benefits bother to be actively seeking work”
You gosh.
Try to get interviews from November through New Years or in July and August for that matter. Mentally closed for the holidays or closed for vacation.
It takes two to create a job search. Most folks who have never been out of work for an extended time don’t realize this. But it doesn’t take more than the cycle of a year for even the most diehard knock-on-everybody’s-door types to realize what times are pure futility.
The BLS is a poor messenger. They are outdated in their releases.
‘This is a piss poor jobs report.’
Fed-watching coupon clippers — the notorious bond vigilantes — quite agree. The 10-year Treasury yield fell to 2.86 percent Friday from 2.97 percent late Thursday, producing a one-day price gain of 0.875% in the on-the-run T-note of 11/15/2023. That’s nearly four months worth of interest, overnight! Or a lotta cat food, if you live on them coupons …
It suggests that after flirting repeatedly with the round-number 3.00% level, yields might have risen too far, too fast. Three percent is now a tempered-glass ceiling — don’t break it!
Washingtonsblog link to the Fed predicting that our economic output in future (no real timeframe?) will be 7% less than it was when? 2006? This looks like the magic 6% solution to maintain unemployment to keep a lid on an overheating economy to control inflation. With an added 1% to mitigate all the QE? So even with a declining population and a retiring boomer generation by around 1% per year, the Fed has upped the ante for the amount of austerity that will be necessary to control inflation going forward? The 7% solution? Does this translate into a directly correlated 7% unemployment rate? Kinda makes sense that the Fed first used 7% as its new benchmark then. But when it got caught out manipulating the BLS stats it had to equivocate. I’d like to know what this 7% austerity is in real unemployment terms since productivity doesn’t seem to be going down. It looks like the Fed is seeking a real unemployment figure of 12 to 17%. And part of keeping inflation down is keeping interest rates down, right? Whatever.
I am not sure it was “that” bad of report when the previous reports since August were rip roaring good. Look at household data employment, rocketed it did.
December looked like a bit of a pause. Big deal. Then January resumes and people wiggle their noise. Got to be careful with the BLS. They are 20 years behind the times and still release 20th century smokestack employment data called the “NFP”. That should be abolished. It no longer captures modern post-industrial job growth correctly and that is being seen in U-rates.
Seasonally adjusted from the Household survey
in August employment decreased 106,000
in September it increased 91,000
in October it fell 785,000 due to the government shutdown
in November it grew 958,000 due to the rebound from the government reopening
and in December it increased 143,000.
I agree with your point about outmoded categories and methodologies. The methods they are using simply don’t fit a de-industrializing economy filled with extreme wealth inequality and ongoing economic crises. The Household and Business surveys have wildly different levels of accuracy and measure somewhat different populations in somewhat different ways. The result is that while the Business survey is more accurate the Household survey often is more informative. The nature of work needs a serious rethink both in terms of job quality and societal usefulness. A worker works one hour at minimum gets the same weight as a full time worker making a decent living. A CEO who is looting the company and making billions to the great detriment of society is counted just the same as that full time worker doing something actually positive for society, while at the same time someone taking care of a sick relative or parent isn’t counted at all. Baby Boomers are retiring and younger people can’t find work. The first of these has been known that it was coming for years. Yet nothing was put into place to track how this would affect the labor force. We are six years out from the recession which began in December 2007, but it continues to be treated like any other recession, indeed as past history, even though the only similar recession was the Great Depression and the effects of the 2007 recession and 2008 meltdown continue to shape the economy and the labor force. As a result of all this, almost nothing that gets reported can be taken at face value. It all has to be recast and recalculated to come up with something that approximates what we actually see going on in the economy.
I think this might have more to do with the seasonal adjustment data. Remember for the last few years, the January/February timeframe would come in way above expectations. Some people suspected that this had something to do with seasonal comparisions of 2008/2009, which threw the data series out of whack. Because, then, sometime in the spring, job numbers would miraculously return to normal after beating expecatations in the beginning of the year, by a wide margin.
So I suspect that now the seasonal data no longer includes 2008/2009, as it has aged off. I’m surprised that nobody seems to have picked up on this.
“Spokescritter”? I prefer “spokeshole.” Fixed it for ya.
I would prefer to see mean rather than average hourly and weekly earnings. Those making millions of dollars per hour must be severely skewing results.
One way to look at this is that production and nonsupervisory workers make up about 4/5 of the labor force. Knowing this and the average for all workers, you can calculate what the top 20% are making on average too. Here we are talking wages, not overall compensation. It’s important to remember that CEOs for large companies can make most of their money from stock options and they also have other tricks like deferred compensation.
I had a thought while driving today. The manufacturing workforce in the US bears the same relationship to Chinese manufacturing that the landless whites (the majority of the white population in the ante-bellum world) bore to the slave economy in the 19th century. They can’t buy a job, just like so many 19th century white Southern workers. Cotton was the China of the 19th century. US cotton exports grew by 7% a year from 1800 to 1850. Not too far from China’s GDP numbers for the last thirty years.
Thanks for the report Hugh. I do appreciate your analysis and agree with your observations. I think that the days of jobs for all is long gone and we need to redefine social contribution and responsibility completely.
In your spare time, how about debunking the CPI numbers on a regular basis? (grin)
I have a feeling we are on the edge of hyper inflation that will not be accurately reflected in the CPI. All of us fixed income folks will be hammered by the results as adjustments will not keep up with cost of living.
there’s been an ongoing discrepancy between the unadjusted employed number from the two surveys since July…from July to December, the unadjusted count of the employed from the household survey fell by 690,000, from 145113000 in July, to 144423000 in December…over the same time frame, the unadjusted non-farm payrolls rose by 2,176,000, from 135577000 in July to 137753000 in December…
here are payroll jobs monthly (000’s):
2013-07-01 135577
2013-08-01 136002
2013-09-01 136612
2013-10-01 137523
2013-11-01 137999
2013-12-01 137753
here’s the raw unadjusted count of the employed from the household survey (000’s):
2013-07-01 145113
2013-08-01 144509
2013-09-01 144651
2013-10-01 144144
2013-11-01 144775
2013-12-01 144423
here’s what the two look like next to each other on a chart:
http://research.stlouisfed.org/fred2/graph/?graph_id=144994&category_id=0
background details, for the record:
change in payroll jobs vs change in employed differs by nearly 3 million since July
maybe they’ll revise it all away in January…