When, oh when, might we finally get to full employment?

September 10th, 2015 at 9:31 am

[This one’s by me and Ben Spielberg]

Source: BLS, Levin, our calculations.

Source: BLS, Levin, our calculations.

Here are two facts: 1) we’ve been adding jobs at an average rate of about 220,000 a month in recent months and 2) we’re not at full employment, by which I mean a very tight matchup between the number of job seekers and the number of jobs.

So, a natural question growing out of these facts would be: when might we get to full employment, assuming we keep up the pace of payroll growth.

We answer that question in the figure above. Here’s the recipe:

First, we use Andy Levin’s total employment gap, which accounts for unemployment, the “missing workforce” (the number of people out of the labor force who’d likely come back in with stronger labor demand), and the underemployed (those working fewer hours than they’d like). That gap stands at about 3.6 million FTE’s right now (full-time-equivalents).

Second, we use the historical relationship between payrolls and FTEs to get a comparable payroll number. Since there are a bunch of part-time jobs out there, we estimate that one FTE equals about 1.08 payroll jobs. Then the miracle of long division occurs and we’re done (1.08*3.6/number of payroll jobs).

Suppose something really bad happens and the pace of job gains gets cut by half. Maybe the Fed goes nuts and doesn’t just tap the breaks but slams them. Maybe we default on the national debt, or some financial bubble implodes. That would push out the time-to-full-employment until at least Dec. 2018, which would mean about a decade of slack labor markets.

At the current pace, we’re potentially looking at the spring of 2017, and if job growth were to significantly accelerate, we could be talking about the end of next year.

Obviously, there are different ways to cut this, but this approach seems the most realistic to us because it deals with various sources of labor market slack. Were one to merely reference the unemployment rate, you’d incorrectly conclude we were already at full employment, and if you were a Fed governor, that might lead you to mistakenly raise interest rates.

Under any of these scenarios, we don’t reach full employment until at least year seven of this expansion, and more likely year eight. That, in a nutshell, is the failure of macroeconomic policy both here and abroad. How to fix it is the subject for another day, not to mention a recent book you might find useful.

Print Friendly, PDF & Email

5 comments in reply to "When, oh when, might we finally get to full employment?"

  1. dwb says:

    “Then the miracle of long division occurs and we’re done (1.08*3.6/number of payroll jobs).”
    No: There are natural new entrants to the labor force due to demographics. Probably 100k or so per month. So, more like (1.08*3.6/(number of payroll jobs-100k))

    • Jared Bernstein says:

      That’s how we originally were doing this! But the labor force has been so jumpy lately we didn’t think we could reliably estimate it.

      • Veddiculture says:

        It hasn’t been jumpy, just aging and declining due to the aging. Financial crisises don’t have fast recoveries. People still amaze me when they think the 1980’s was fast. It wasn’t really until late 87 that full employment really happened.

  2. Veddiculture says:

    Full Employment is unknowable. The Government also has lags in wage reporting data.

  3. Omari West says:

    Interesting post. Fed VC Fischer – no fan of loose money — alluded to this miscalculated employment in his comments at Jackson hole. I think we can safely say that the demand for employment (perhaps demand for income) still greatly outpaces the supply. Oil price declines would seem to be a mitigating factor (as they inputs to production), while the strong dollar would exacerbate unemployment by dampening exports. However, I think the Fed has long ago reached the limits of what it can do with monetary policy. Time for some real public policy shifts – and that’s where the political system has failed to deliver over the duration of the recovery.