Following on from my column last Friday on the methodology behind the Bureau of Labor Statistics' Monthly Employment Report -- and just in time for February's report, which comes out tomorrow at 8:30 a.m. ET -- let's look at the numbers, themselves.
As I mentioned in my previous column, the employment report is actually composed of two surveys: the establishment survey (CES), and the household survey (CPS). Therefore, it may be unrealistic to expect the two surveys to agree.
But it really shouldn't be. If employers (surveyed in CES) are hiring more workers (surveyed in CPS) the numbers should converge. Each job filled should match each job taken. But the numbers don't always show that, and for a numbers geek like myself, that is frustrating, to say the least.
Let's look at the most recent data from January, which was reported Feb. 5.
The most widely quoted figure is the "jobs number," the month-on-month change in nonfarm payrolls from the CES. That figure, as you no doubt read, was 151,000 for January. It is seasonally adjusted, as are most government data.
On a non-seasonally adjusted basis, the number of people employed in nonfarm jobs fell by nearly 3 million (2,989,000 to be exact) from December to January and registered 141.1 million as opposed to the seasonally-adjusted figure of 143.3 million.
So did we lose 3 million jobs or gain 151,000? And what about the 2.2 million jobs that exist in the seasonally-adjusted world but not in the as-reported world?
Well, the difference is driven by the holidays and the seasonal demand for labor. That's because everyone's Christmas shopping involves piling into Dad's Oldsmobile Six next to an annoying, heavily-bundled little brother and heading to the big department store downtown. Actually that's A Christmas Story, and the movie was set in the 1940s. Things are different now, thankfully.
Companies like Amazon (AMZN) hire seasonal employees, and UPS (UPS) hires employees to keep up with Amazon, but the reason Amazon is a $270 billion market-cap company is that clicks are much less labor-intensive than bricks.
So, I'm highly skeptical of those adjusted numbers, and if you want to read more on that topic hop on to Google and check out the excellent work former OMB Director David Stockman has done on the BLS' seasonal adjustments. He's not a big fan, either.
The household survey is even more frustrating. The metrics are basically top-down. The BLS surveys a ridiculously small sample -- see my prior column for the numbers -- and from that determines:
- Total Population
- Eligible Civilian Labor Force
- Total Employed
- Not in Labor Force
The math is simple. Labor force over population gives the participation rate; total employed over population gives the employment to population ratio; and unemployed over labor force gives the oft-quoted unemployment rate.
But these numbers suffer from the same strong seasonal biases (and large future-month revisions) as the employment survey. In January the not-seasonally-adjusted employed population was 149.0 million, down from 149.7 million. On a seasonally-adjusted basis however, the number of employed grew by 515,000 to 150.5 million.
So, what is the total number of employed persons in this country? The two surveys' key numbers differ by about 8 million people. That's kind of an important amount, and I don't think that farm workers make up all the difference.
The upshot is that because fewer people are in the labor force, the percentage of unemployed has a natural downward bias - and the unemployment rate of 4.9% is totally meaningless.
In contrast, as of January, 95.1 million people were reported as "Not in the Labor force" as per CPS data. That's 37.7% of the population and that figure is much higher than it was 10 or 20 years ago. What are those people doing? Are they all driving Uber cars and working as social media consultants?
I have no idea, but I care much more about that 95 million than the approximately 8 million who show up as "unemployed." The results from Super Tuesday indicate to me that those "unaccounted for" 95 million are unhappy at best, and, in many cases, just plain pissed off.
Forget the CES/CPS alphabet soup and pay attention to the forgotten folks. They are making themselves heard.