It’s widely understood that the pace at which the US job market gets back to full employment—assuming it does so before the next downturn—is in part a function of the size of the labor force. A larger labor force, the more jobs it will take per month to lower the jobless rate.
This a germane observation given that last month’s large decline in unemployment, from 6.7 to 6.3 percent, was wholly driven by an equally large decline in the lfpr (labor force participation rate). As the WSJ pointed out last Friday, assuming the lfpr stays at its current post-recession low of 62.8%, if payrolls continue to grow at their average pace over the past three months—about 240K/month—the job market would hit the top of the Fed’s full employment unemployment rate of 5.6% in just six months (let’s label this monthly value MTFE: “months to full employment”).
Sounds good, right? But—and this is no critique of the WSJ piece, which noted this point—that estimate is likely biased down due to the depressed lfpr.
The WSJ used this handy Atlanta Fed calculator to come up with the six month estimate and I’ll do the same to make the lfpr point. The calculator allows you to plug in various assumptions and then spits out the number of payroll jobs we’d need to create per month to hit your specified unemployment rate. My point is to look at how the MTFE increases changes when you plug in what I think is a more realistic lfpr.
If you take a three-month moving average of the lfpr—to smooth out the monthly bips and bops—you’ll find it’s down about three percentage points from its pre-recession peak, roughly 66% down to 63%. There’s been considerable argumentation as to how much of that decline is structural, as in people leaving the labor force for good, say for retirement, and how much is cyclical, such that they’d come back once the job market tightens up some of higher wages beckon them back. I review the arguments here, but let’s take what I’d say is a fairly conservative view and split the difference.
So, how does that six-month figure change when we add 1.5 points to the lfpr (i.e., set it at 64.3%) but keep everything else the same (i.e., still assume that full employment equals 5.6% unemployment)? It goes up to a few months short of three years (see figure), well beyond the six months you get if you stick with the current lfpr.
If you then assume, as I’d argue you should, that full employment is closer to the lower end of the Fed’s range, 5.2% (and true full employment may well be lower than that), then it would take another five months to get there.
The current expansion is already middle-aged; another 40 or so months of growth would make it among the longest in the postwar period. That may well occur, but there are two points to keep in mind here: first, a partial return to a more normal lfpr amidst steady payroll gains at their recent pace will significantly extend the time it takes to get back to full employment, and second, these weak, jobless recoveries, characterized in this latest case by severe policy neglect under the guise of austere fiscal policy, make it much harder to achieve full employment before the next recession.
One other point can be drawn from the Fed calculator that some of my readers will find corroborating (I’m thinking of those who worry about the impact of increased immigration in demand constrained periods, for example). Economists who talk and think about full employment typically and mostly do so from the demand side. But the pace of supply of course matters too and you can fool around with the calculator in this regard by juicing the rate at which the working-age population grows.
For example, let’s say that we take the population growth rate up from its current monthly moving average of 0.077% to its pre-recession rate of around 0.1%. That too significantly lengthens the TTFE to a bit over four more years. But basic macro dictates that faster growing population also increases demand, and we can incorporate this idea by boosting the job creation number, say from 240K per month to 280K, which brings the TTFE back down to three years.
Any way you cut it, even assuming the recent trend in payrolls sticks, unless you also accept that the labor force remains stuck in the basement and full employment is 5.6%, MTFE is a lot longer than six months.
Source: Atlanta Fed Calculator; jobs per month target held at ~240K except in last scenario.
1) WSJ estimate (I get 7 mos, not 6, btw, but close enough)
2) Add 1.5 percentage points to current LFPR (62.8% to 64.3%)
3) Full employment unemployment rate 5.2% instead of 5.6%
4) Increase working-age population growth rate from 0.077% to 0.1%
5) Move jobs target from ~240K to ~280K.