The Bureau of Labor Statistics (BLS) February, 2012 Employment Situation report is out. The “official” measure of the nation’s unemployment rate is unchanged at 8.3%. You can read it here. Now the blogosphere will delve into the numbers and issue their interpretations as to their meaning. We’ve witnessed the generally steady decline in the unemployment rate for months now and the statistics have revealed the reason for the drop has been those dropping out of the labor force and no longer counted rather than actual job creation.

The February 2012 report is different from the recent pattern. The January 2012 report showed that 1.177 million people were added to the “not in labor force” category. The February 2012 report shows a decrease of 310,000 in the same category. That’s quite a turnaround. Let’s look closer.

From December 2011 to January 2012, we added 284,000 non-farm payroll employees, had 1.177 million drop out of the labor force and the unemployment rate dropped from 8.5% to 8.3%. From January 2012 to February 2012, we added 227,000 non-farm payroll employees, had 310,000 fewer people drop out of the labor force and the unemployment rate stayed the same. Hmmm. We created 57,000 less new non-farm payroll employees than the month before. We had a swing in the “not in labor force” category of 1.487 million people. Yet the rate was unchanged. Confusing?

BLS methodology always is. Here’s some further confusion. The BLS Establishment survey shows an increase of the civilian non-institutional population of 3.584 million. What is this statistic? From the BLS.

The civilian noninstitutional population consists of persons 16 years of age and older residing in the 50 States and the District of Columbia who are not inmates of institutions (for example, penal and mental facilities and homes for the aged) and who are not on active duty in the Armed Forces

To be clear, this is not the overall population increase of the country, only the people included above. Now, you can’t get monthly revisions to the actual population numbers from the U.S. Census Bureau, so the BLS institutes a model in its methodology to estimate these numbers. For a comparison, consider the Federal Reserve of St. Louis (FRED). Take a look at this chart of their estimate of the TOTAL population of the country, not just those of working age. If you’ll scroll down, you can see their source is also the U.S. Census Bureau. If you look to the left of the chart, you can see the recent month-by-month changes.

I looked at their data and saw an increase of the total U.S. population from December of 2010 at 310,760,000 people to December of 2011 at 313,020,000. The difference is 2.26 million people. How can this be? The BLS and FRED both use U.S. Census Bureau numbers, yet the BLS says the civilian non-institutional population alone expanded by 3.54 million in the last year and FRED says the entire population increased 2.26 million. A difference of 1.28 million people. In summary, the BLS says the population is expanding much faster based upon their methodology than other Federal entities utilizing the same source data.

What to take away from this? The same thing I say over and over. The data is unreliable regardless of whether it’s positive or negative. BLS methodology uses modeling and surveys extensively to issue reports that determines policy and, in some cases, elections. That’s a serious concern. Anyone should be able to extract numbers from these reports to put their candidate or party in a positive light. The end result is invalidating the BLS entirely.


4 thoughts on “February unemployment rate unchanged – should we even pay attention?

  1. Hi, Spellcheck, I just came upon your blog today and I’m going to see if I can help you out a bit. I write a blog that concentrates on employment numbers, and I try to explain these things as simply as I can. First of all, I would suggest that, as you don’t quite understand the data, you refrain from making any decisions as to the validity of BLS/Census data until you understand what you are looking at.

    I’m just going to take the items that you’ve discussed one at a time and I’m going to see if I can help you out.

    So here’s your first statement:

    “We’ve witnessed the generally steady decline in the unemployment rate for months now and the statistics have revealed the reason for the drop has been those dropping out of the labor force and no longer counted rather than actual job creation.”

    May I ask you what statistics have revealed this and where they are?

    First of all, you are right that the number of people “not in the labor force” is going up, but there is real ongoing job growth.

    The BLS establishment report (not the CPS population report which is the report that shows the unemployment rate) shows us that private employers have created 3,939,000 jobs in seasonally adjusted numbers over the past two years. The independent ADP report which surveys private employers also comes up with an estimate of the number of private jobs added in seasonally adjusted numbers over the past two years. Their number is 3,502,000. ADP’s report for February can be found here:

    It contains a graph which correlates the ADP job estimates with the BLS estimates, and, considering the size of the population, their numbers are surprisingly close. Remember these two organizations use different data sources, even though both have a “sample” and apply varioius statistical techniques to that sample to get their numbers.

    The low number of private jobs as reported by BLS in early 2010 was 106,773; the low number of private jobs reported by ADP was 106,747. BLS now estimates 110,711 and ADP now estimates 110,249 private sector jobs. I’d bet on BLS as these numbers are revised over the coming months as more specific data is available. I can explain those revisions as well if you like.

    So, yes, two separate sources, one from the big bad old guv’ment and one from private-sector payroll processor ADP, both estimate that we have added millions of jobs in the past two years.

    People leaving the labor force (or just bad estimates that are readjusted) or not, we are adding many, many jobs.

    Anyway, that’s just a start.

    You sound like an intelligent person, so I’ll be happy to come back and explain more.

  2. Sounds like you’re quite happy with the employment situation as is. If your concern is correcting bloggers interpretation of statistics, I would encourage you to visit this site and take up your concern with the author-

  3. I have looked at Mish’s numbers and I have considered his concerns before. I have already addressed many of Mish’s issues in my blog. He brings up many points that do require explaining, but the BLS actually explains most of those points very well. I’m not sure why Mish and others are having a hard time understanding what the BLS is doing and why it does it.

    For example, the large increase in the civilian population in January 2012, resulting in a large increase in the “not in the labor force” bucket, is the result of adjustments after the 2010 Census. In short, the Census estimates the population, and the breakdown of the population by age, sex, ethnic background, etc., monthly. After the the decennial Census is completed, the Census adjusts its population figures accordingly.

    In the case of the 2010 Census, the estimates of the “civilian non-institutionalized population 16 and over” were off by 1,700,000, or about 12,000 people a month each month over the past ten years. Not only that, but most of those underestimated people were in the 55+ age bucket. 1,500,000 people were underestimated in the 55+ group over those ten years. From monthly sampling, we know that the labor participation rate for people who are 55+ is only 40%. So, of those underestimated people, 900,000 are not going to be in the labor force, and those 900,000 are going to go right into the “not in the labor force” bucket. The 55+ “not in the labor force” did increase by 985,000 people between December and January.

    Meanwhile, the 25-54 age group was OVERestimated in monthly Census estimates. Those people have a high labor force participation rate, so, again, the adjustments wound up decreasing the number of people in the 25-54 age group and therefore decreasing the number of people in the labor force.

    I’m still not sure I’m explaining this well, but that’s just a start towards explaining why the civilian non-institutional population was adjusted upwards by 1,700,000 people in January and the “not in the labor force” bucket received most of those undercounted people. a7utr.

  4. I would be much happier with the employment numbers if the Republicans weren’t intent on screwing the American people with their constant obstruction for political gains. The biggest sector that is losing jobs is in the “local government-education” bucket which has lost over 200,000 jobs over the past two years. How can that be a good thing.. unless you are a Republican who wants to privatize education and turn our kids into somebody’s profit center?

Comments are now closed.