For whatever reason, political or otherwise, I often hear people insist that the unemployment problem is subsiding.
“The economy may be bad, but at least we’ve gotten better.”
“Look, the unemployment rate used to be 10 percent, but now it’s 7.4 percent! We’re moving in the right direction.”
But are we really?
To answer this question accurately, I’ve been spending some time examining macroeconomic data aggregates on FRED. And now I’m going to share with you what I found.
Here’s that headline unemployment rate (blue), with the “natural” unemployment rate (red). As you can see, U3 has been declining:
(Update: For those of you less familiar with macroeconomics, the red line approximates a “natural” rate of unemployment in a healthy economy.)
Furthermore, monthly job growth has been in positive territory:
So the economy really is digging its way out of this hole, right? Not so fast.
One important factor that must not be overlooked is population growth. Just eyeballing this next chart, we can see that the total US population is netting about 160,000-220,000 new people every month:
But this chart shows growth in the total population. In reality, not everyone in the entire population is prone to work, or even falls under the working-age category. So the relevant population would be those who are of working age (15-64). Even still, only about two-thirds of the people in that category are typically employed.
To obtain a graph that thoroughly accounts for these factors, I did several things. I found a population metric for everyone ages 15-64, and I found the metric for total payrolls and subtracted out those who held jobs at 65 and older. I scaled down the population data (red) to about 67 percent of the actual, in order to account for the longer-term nonparticipation trend and align it with the payroll (blue) at roughly the natural rate of (un)employment. Here is what I got:
As you can see, the Great Recession is characterized by a large and persistent employment gap. However, it does seem that the gap has slightly contracted. In 2010, it was around ten million; in 2013, that number is about seven million. That means we’ve had some modestly helpful job growth, right?
Well, that’s not the whole picture.
Before I even factored out those who are 65 and over, I had made a similar chart juxtaposing payrolls and population for everyone age 16 and above:
The two trend lines here run much more parallel. This goes to show that much of the gap decrease in the 15-64 chart is through attrition of the working-age population as the earliest baby boomers turn 65. We can observe this in the next chart, which shows the semiannual growth rates of payroll (blue) and working-age population (red):
But there’s more. One of the biggest factors behind the illusory decline in headline unemployment is a steadily decreasing labor force participation rate as jobless workers have become discouraged. To remove the effect of nonparticipation from retirement, I narrowed this next chart down to the ages of 25-54.
Also helpful is the employment-population ratio, which is basically the number of people employed divided by the population:
That by itself should be enough to show that employment has not significantly improved. By syncing the employment-population ratio with the labor participation rate, this final chart shows precisely what is misleading about that original headline unemployment chart:
Well, there you have it. The Great Recession has had a very slow and protracted “recovery,” albeit with a declining headline unemployment rate. After poring over the relevant data, it seems the dropping unemployment rate is mostly an illusion, as the post-recession economy has been characterized by: (1) job growth that is only slightly above population growth, (2) stepped-up rates of retirement from the maturing of baby boomers, and most significantly (3) decreased labor force participation rates among the core working-age population, as many of the unemployed are discouraged and have ceased looking.
So, has unemployment fallen significantly? Answer: It really hasn’t.
Update: There is probably a secular downward trend in labor force participation that predates the Great Recession. Including it might make the full-employment threshold look something like this:
Ostensibly, this reduces the cyclical component of the labor participation decline (and its contribution to declining headline unemployment), albeit not by very much.