Reality Check on Employment Numbers

On Friday, the big news was that the unemployment rate had fallen below 6 percent for the first time in six years. That’s swell news, but the headline unemployment rate tells us virtually nothing at all about the real employment situation. Since the unemployment rate is a function of both the labor force size and the total number of people who self-report as employed, we need to get a sense of both labor force dynamics and total employment. Below, I’m going to use the government’s own numbers, so keep in mind this data is the rosiest picture that BLS could credibly paint. I will not be using the so-called “seasonally adjusted” (SA) numbers, because they’re more heavily manipulated than the not-seasonally adjusted (NSA) numbers. Also, seasonal adjustment is simply unnecessary and adds  totally unnecessary complexity to the data. Everything I look at below is not adjusted, and is from the BLS.

Here’s the unemployment rate graph:


The unemployment rate bottomed out near 4 percent during the last expansion (2002-2008), and it then shot up to over 10 percent. Ever since then, it’s been heading down steadily. The headline SA number for September 2014 was 5.9 percent, and in this case the NSA number was even lower, at 5.7 percent for September. That’s down from a year earlier, when the rate was 7.0 percent.

How great! What a big drop. We’re finally back to what was ten years ago considered a mediocre unemployment rate. But it’s better than ten percent, right?

Sure, it’s better, but what really matters, in terms of the unemployment-stats game is how much actual employment opportunity there is. The fact of the matter is, the total number of employed persons in the US has gone basically nowhere since 2007.  We see this if we look at the components of the unemployment rate, which are the labor force size and the total number of employment persons. This is the Household Survey, which means that the data is based on surveys of actual persons who are asked if they want to be employed, and if they have actually managed to find employment. People who have given up looking for work, or gone back to school, or just accepted a lower standard of living because they’re premature retirees who’ve given up permanently, are all excluded from the labor force. So, looking at the components of the unemployment rate, we see:


When the gap between the two lines is big, the unemployment rate is high. And when the gap is small, the unemployment rate is smaller. It’s perfectly possible to get a decline in the unemployment rate without adding any new jobs at all. All you need to have is a decline in the size of the labor force. We saw this at work back in 2010 and 2011, when the unemployment rate was dropping, but employment growth was near-stagnant. This was because the labor force size fell slightly, in spite of the fact that 3.3 million people continued to graduate from high school each year. Did 100% of them go to college full time? And what happened to all the people who graduated from college in those years? Well, they apparently didn’t enter the labor force. Labor force growth continues to be anemic, and it’s certainly not just because baby boomers are retiring. Many people are simply electing to leave the work force for a variety of reasons, including declines in opportunities for work. Many critics of the administration have correctly pointed to the fact that labor force participation is at very low levels.

Meanwhile, if we look at the line for the total number of employed persons, we find that the number is only now just starting to get back to its old peak level, earlier achieved in 2007. Since there are real seasonal differences in employment totals, I have broken out employment totals by month so we’re comparing apples to apples. So, look at September below which includes the most recent data. Since the scale does not begin at zero, the burgundy bar looks like it’s outpaced the former peak level (the light blue bar) for September (2007) by a small but significant amount, but in fact the difference is excruciatingly small. Over that seven year period, from 2007 to 2014, the total number of employed persons, has increased by 520,000 people. That’s an increase of 0.35 percent over that whole period. So, with a few million people graduating from high school and college each year, every year, the US has managed to add half only of a million jobs in that time.


We might also note that the Household Survey tells us how many people are employed, but not how many people are underemployed, or working two jobs to make ends meets. In fact, people looking to present a rosier picture of job growth will turn to the other measure of employment, the Establishment Survey, to find larger job growth. The Establishment Survey looks at total payroll jobs, not at employed persons. So, a person who has two jobs will sometimes show up as two payroll jobs in the establishment survey. Indeed, Obamacare has likely helped push up totals in this survey: aA employers cut back on hours to avoid Obamacare mandates, and employees are forced to find a second job to pick up the slack in hours, this will add to the payroll job numbers. There are other factors as well, of course, but we do indeed find that the job growth looks better in the Establishment Survey, although not that much better:


Overall, employment in the establishment survey has also just barely gotten back to where it was in 2007, but  comparing September 2014 to September 2007, we find that, percentage-wise, payroll employment is up by twice as much as the Household Survey. It’s up 0.98 percent! (It was up only 0.35 percent in the Household Survey). If broken out by month to account for seasonality we see that September 2007 to September 2014 is nothing to write home about:

emp2Now, most news outlets tend to compare year-over-year, not peak to peak. And, if we look at YOY numbers, they do look a lot better. That jump from 2013 to 2014 in the graph looks pretty big, and as with the Household Survey, the -year-over-year numbers of the Establishment Survey look pretty good. From Sept 2013 to Sept 2014, payroll jobs increased 1.9 percent or 2.6 million jobs. But that ignores the fact that we’re really still in “replacement” mode when it comes to job creation. The 2009-2010 hole in employment has just now been exited, and during all that time, millions of people lost income, experience, and the ability to save and invest for the future.

It’s great that the data looks relatively good compared to the dismal  numbers of the past seven years, and to last year, but it’s hardly time to throw a party when we consider what’s been going on for seven years, and the real social costs that the Fed-induced crash of 2008 has left in its wake. The job numbers have just now turned positive, so we have to ask ourselves if job creation can possibly survive any cutbacks in monetary pumping. The job growth comes on the heels of multiple years of quantitative easing. The Fed has been hoping and praying since QE1 that the economy would just magically revive itself with some loose money, but even after several rounds, employment is now just back to pre-recession levels. What would happen if interest rates went up and/or the true money supply began to decline? It doesn’t look promising.


3:45 pm on October 7, 2014