This from Michael Becker explaining how the unemployment number we watch so closely is calculated. We all suspect this number can be jiggled. This explains how.
How is unemployment calculated? Excellent question, and there are several answers. Hey, it’s the government and you should know by now that NOTHING they do is simple or straightforward. Here are the basic terms we need to understand:
- Civilian non-institutional population (referred to below as “population”): the adult population of the US, the incarcerated, those who are disabled in a home for the aged, or on active duty military service. This is an estimate based on census data.
- Employment: all persons who did any work at all as paid employees, were self employed or worked in a family business, or temporarily absent from a regular position.
- Unemployment: persons who had no employment and who made specific efforts to find work.
- Civilian labor force: #2 plus #3.
- If you are unemployed and are not collecting unemployment you are not counted as “looking for work” no matter how badly you’d like a job.
- If you’ve gone back to school because you can’t find a job and you can bleed student loans for living expenses, you’re not counted.
- If you’ve retired because you couldn’t find work but would really like to have a job so you can eat more than four meals a week, you’re not counted.
- Get the idea?
- Workforce participation: an estimate of the percentage of #1 who are working or looking for work (#4).
Why 8.3% is really 11.5%. And a tip of the hat to some of the smartest guys around, the folks at ZeroHedge, who crunched the numbers. If you’re really interested in learning about how public finance works, not just in the US but all over the world, they are an indispensible resource and you should be at least checking in over there two or three times a week.
In January, the population was 242,300,000. That represents an increase of 1,700,000 people over the December population.
The BLS reported a seasonally adjusted labor participation rate of 63.7% and when factored against the population, that makes the implied workforce 154,400,000. In reality, participation is more like the long term average of 65.8%, because old fogeys like me not only aren’t retiring at the rates BLS assumes, we are either working overtime or have a second job (you’re reading mine, and thank you very much). The difference in implied vs. real participation adds 5,000,000 to people to the workforce, making the real workforce 159,400,000. The BLS and the Obama Administration in general are purposefully choosing to ignore these workers.
The BLS does account for 12,800,000 unemployed workers and they base their calculation of the unemployment rate on that number. Note that they are purposefully missing almost one-third of workers who are not employed but would certainly like to have a job. By adding those workers back into the unemployed figure, the actual number of unemployed is 17,800,000
If we divide the real number of unemployed workers by the BLS reported workforce, the actual unemployment rate is NOT 8.2%, it is 11.5%.
Take a hard look at the following chart that shows labor force participation by age group and you’ll see that fogeys are indeed working longer and harder. I don’t expect this trend to slow down, at least in what’s left of my lifetime.
Further indication, again provided by our friends at ZeroHedge in an analysis of actual employment numbers for January, note that there are 7.4 million fewer jobs than there were four years ago. Those were housing and credit bubble jobs that will never come back. That number lends credence to the analysis above, relating to the five million people the BLS chooses to no longer count in the workforce, because they’ve been out of work so long.
The above analysis also notes that there was a big bump in payroll tax collections that happened at the end of December, and continued through mid-January that is unexplained by any of the employment data. That spike dropped last week and, it doesn’t look like it will continue into February.
The bottom line here, in slightly more detail than I first noted, is that BLS is reporting seasonally adjusted numbers, and because of some unexplained spikes in the early part of the month that have since bottomed out. The February data should be interesting, and frankly I’m expecting the sub-heads on next month’s stories to include the word “unexpected” once again.
–Michael Becker
