If you have "no place to go," come here!

BLS Jobs Report Covering February 2012: I Wish the Numbers Added Up

The BLS jobs report covering February 2012 showed not only a big gain in jobs for the month (227,000) but also revisions upward for both December (203,000 > 223,000) and January (243,000 > 284,000). The unemployment rate, however, remained unchanged at 8.3% reflecting a return to the labor force of some of those the BLS had chosen to define out of it.

So I suppose the important question is how much of this is real. Offhand I would say it is a mixture of stimulative early in the year government spending (remember its fiscal year begins in October) moving into the economy and modeling and margin of error effects. But let's look at the numbers.

The potential labor force as represented by the Civilian Non-institutional Population over 16 (NIP) increased by 166,000 from 242.269 million to 242.435 million. The seasonally adjusted employment-population rate was 58.6% (an increase of 0.1%). This means .586(166,000) or 97,000 were needed for the month just to keep up with population growth. So even if we subtract this out, and bearing in mind that employment and jobs numbers come from two different surveys, the Household and Establishment or business surveys, respectively, that are only roughly comparable, the economy beat population growth by 130,000.

For the Household data, I will give the unadjusted numbers followed by the seasonally adjusted ones. Unadjusted the labor force grew by 629,000 in February from 153.485 million to 154.114 million. Employment increased by 740,000 from 139.944 million to 140.684 million and unemployment declined 111,000 from 13.541 million to 13.430 million.

740,000 – 111,000 = 629,000

Seasonally adjusted, the labor force increased 476,000 from 154.395 million to 154.871 million. Employment increased 428,000 from 141.637 to 142.065 million. Unemployment increased 48,000 from 12.758 million to 12.806 million.

428,000 + 48,000 = 476,000

The unadjusted unemployment rate decreased a tenth of a percent to 8.7% and the seasonally adjusted rate remained unchanged at 8.3%. More importantly, the unadjusted the participation rate increased two tenths of a percent from 63.4% to 63.6% as did the seasonally adjusted rate 63.7% > 63.9%. This is indicative of both a major influx of job seekers into the labor force and that many of them were successful in finding work.

But at the same time, these unadjusted and seasonally adjusted numbers are 2 and 3 times, respectively, the number of jobs reported created in February. I think this is a case where the 400,000 error margin of the Household survey vs. the 100,000 error margin of the Establishment survey is at work. At least, this month the seasonally adjusted and unadjusted Household numbers, unlike last month, were more nearly aligned in terms of direction and sign.

The BLS measure of its undercount, that is those it defines out of the labor force, is its category Not in Labor Force, Want a Job Now. Unadjusted, this declined 119,000 from 6.495 million to 6.376 million. A subset of these is the marginally attached, those who have looked for work some time in the last year but not in the last month. This subset is one of the 3 components along with the unemployed and involuntary part timers which makes up the broader U-6 measure of un- and under employment. The marginally attached decreased 122,000 from 2.730 million to 2.608 million.

The last component of the U-6 is involuntary part timers, those who work part time but would work full time if they could. These declined 111,000 from 8.230 million to 8.119 million.

Taking the three components together the U-6 represented 23.533 million American workers or 14.9%, down from 15.1% last month.

I calculate alternative measures to the U-3 and U-6 which I call real unemployment and disemployment. Basically, I take a participation rate of 67% indicative of a solid expansion and multiply this by the Non-institutional Population (NIP). This gives me a number of what the size of the labor force should be. Subtracting the size of the current labor force from it gives me the size of what the BLS undercount actually is.

So .67(242.435 million) = 162.431 million

162.431 million - 154.871 million = 7.56 million, down from 7.925 million last month.

Real unemployment = U-3 unemployment + the calculated undercount, this is the number of Americans out of work but who would be working if jobs were available to them

12.806 million + 7.56 million = 20.366 million.

This is down 317,000 from 20.683 million last month and yields a real unemployment rate of 12.5%, down from 12.7% last month.

Disemployment = U-3 unemployment + the calculated undercount + involuntary part timers

12.806 million + 7.56 million + 8.119 million = 28.485 million.

This is down from 28.913 million last month and is equivalent to a disemployment rate of 17.5%, down from 17.8% last month.

With regard to race, white unemployment edged down a tenth of a percent to 7.3% White adult male unemployment decreased similarly to 6.8%. White adult female unemployment was unchanged at 6.8%. White teen unemployment increased from 21.1% to 21.3%.

African American unemployment after a big drop last month increased from 13.6% to 14.1%. The principal driver here was among African American adult males whose unemployment rate rose from 12.7% last month to 14.3%. Unemployment among African American adult females was relatively unchanged at 12.4%, down from 12.6% in January. African American teen unemployment declined from 38.5% to 34.7%. African American unemployment still remains at something like twice the level among whites.

42.6% of the unemployed as defined by the BLS had been unemployed for 27 weeks and more (that is more than 6 months). This compares to 42.9% last month and 43.9% a year ago, so really very little near or longer term change.

Turning briefly to the Establishment data, the private sector created 233,000 jobs and government lost 6,000, netting 227,000 jobs. 82,000 came from professional and business services, of which 45,200 were temp positions. Healthcare added 49,000.

Total weekly hours remained unchanged at 34.5. Average hourly earnings for all workers increased 3 cents to $23.31.

So something happened in February but we won't know until the data for May come out, when the upside of seasonal effects are gone, what exactly. It looks like a good month job-wise as were December and January. But each of these had its own problems with the data. December is based on a different Census. In January, the seasonally adjusted and unadjusted numbers were wildly at odds with each other, and in February, we had two to three times the number of people coming into the labor force than the number of jobs created for them. The major difference between the Household survey and the Establishment survey in this regard is that the Household survey counts agricultural workers and the self-employed. Only there was little change in the numbers of either in February. So what we are looking at seems to be a margin of error effect in the Household data. We should have a better grasp on this, as I said, in June when the May data comes out. As for the Establishment data, good news on the job numbers front but no movement on hours or wages. In a true recovery, we would expect to see progress in both.

No votes yet


Alice X's picture
Submitted by Alice X on

Every well considered effort to explain the numbers is appreciated by me. This is clearly one. Thank you!

At this rate we will be to full employment (whatever that is) by 2020, in theory. But with peak fossil energy then long in the rear view mirror we will be well into our final decline.

This economy has long been predicated on cheap energy, waste and more recently, accounting fraud.

There is no way out, but cheers anyway.

Submitted by hipparchia on

The major difference between the Household survey and the Establishment survey in this regard is that the Household survey counts agricultural workers and the self-employed.

i did not know that. thanks for this info.

Submitted by Hugh on

The monthly job reports always have a few pages where they set out their methodology. These are the differences they give between the Household and Establishment surveys.

The household survey includes agricultural
workers, the self-employed, unpaid family
workers, and private household workers among the
employed. These groups are excluded from the
establishment survey.

The household survey includes people on unpaid
leave among the employed. The establishment
survey does not.

The household survey is limited to workers 16
years of age and older. The establishment survey is
not limited by age.

The household survey has no duplication of
individuals, because individuals are counted only
once, even if they hold more than one job. In the
establishment survey, employees working at more
than one job and thus appearing on more than one
payroll are counted separately for each appearance.