Isn’t the Unemployment Rate Just an Inadequate Measure of Slack?

January 17th, 2014 at 4:49 pm

I’ve written pretty copiously of late on how the current unemployment rate is unusually uninformative about the current weakness of the job market.  Due to many potential job seekers leaving the labor market—and the fact that you’re only counted in the jobless rate if you’re looking for work—it’s biased down.  In my estimation, the bias right now amounts to a couple of percentage points, which is historically large.

But these writings have triggered a very good question from many readers: why is unemployment defined this way?  Shouldn’t it be designed to capture those who’ve left the labor force due to discouragement about their job seeking prospects?

The obvious answer is yes, and, in fact, there’s another, more inclusive measure of labor market slack that kind of gets at that issue—it includes “discouraged” former job seekers.  With its clever title of “U-6” this broader measure of labor utilization includes not only discourage job seekers but the partially employed as well, i.e., part-timers who would rather have full-time jobs.  U-6, plotted below along with the official rate, was 13.1% last month, almost twice the 6.7% of the official rate.



FTR, U-6 has fallen even faster than the official rate (4 ppts vs 3.7 ppts compared to their peaks), as both the numbers of involuntary part-timers and discouraged workers have diminished of late.  But is this type of more inclusive measure really better?

I’m not sure.  The official measure has the advantaged of simplicity: you’re unemployed if you’re unsuccessfully seeking work.  Boom.  That’s it.  Everything else involves some degree of psychology.  You say you’re discouraged because there aren’t enough jobs, but how do you know if you’re not looking?  Would the involuntary part-timers really kick up their hours if they could?

Now, these variables–discouraged workers, involuntary pt’ers–are quite cyclical, so I suspect that they’re generally good, reliable measures of slack, that we—and by “we” I mean the Federal Reserve in contemplating whether to keep up their stimulus and the Congress in contemplating the UI extension—should carefully track these days.  The flat EMployment rate is also an important indicator right now.

But here’s the thing: because the labor force has typically held up more than in recent years, we didn’t so much need these other measures.  The good old, very simple unemployment rate was telling us what we needed to know and it was strongly correlated with these other measures such that we wouldn’t have learned much more from them as we did from the official rate.

The problem we face right now is thus less that the unemployment rate is poorly conceived, and more that an unusually large share of the adult population has left the labor force.  The graph below just plots unemployment along with the labor force participation rate with recession shading.  Two points jump out at you.  First, unemployment is much more cyclical than the LFPR, and second, the LFPR is falling faster than it has in its lifetime.


I can make this point about relative cyclical differences between these two variables even more dramatically.  Since a time series can be decomposed into trend, cycle, seasonal, and leftovers (random noise, if you’ve got the rest of it right), I can extract the cyclical component from both series, as I do in the next graph.

The cyclical part of unemployment is in red, and it jumps up and down pretty visibly.  Meanwhile, the cyclical part of the LFPR looks a bit like the EKG of an almost dead person.



In other words, and to both cut to the chase and put this more plainly, the unemployment rate has historically done a good job of reflecting the slack in the labor market because when people lost their jobs, they didn’t give and stop looking, as many have of late.  A big part of that is, of course, the historically very high rate of long-term unemployment, as people just give up after a while.

It also true that some of this labor force decline is older people who would have left soon anyway—this isn’t all about weak demand; there’s a demographic component too.  But besides the immediate policy issues that I’ve been stressing, and stressing out over, like the UI extension, what’s particularly worrisome here is the extent to which persistently weak demand is turning a cyclical problem—weak job growth leading to labor force exit—into a structural one—a permanently smaller labor force and lower potential growth.

This, in turn, means that the cost of bad growth policy, like austerity measures, is even greater than we thought.

Print Friendly, PDF & Email

17 comments in reply to "Isn’t the Unemployment Rate Just an Inadequate Measure of Slack?"

  1. Michael says:

    It reminds me alot of earthquake preparation in California. What I mean is that without major tremors occuring with enough frequency to remind everyone of what is possible, smaller earthquakes tend to create self-fulfilling cycles of complacency and structural inadequacy that isn’t fully revealed until the next major earthquake hits, and then everyone is scratching their head wondering why all the buildings fell down.

    The stable and cyclical nature of the unemployment cycle isn’t going to reveal massive structural changes in the social/political policy right away. The aggregate erosion of a middle-class, factory-based, labor force, like mini-earthquakes hiding the major earthquake building up behind it, is going to give data crunchers alot of semi-false data. Eventually, like a major earthquake, the sudden change in the data isn’t going to make sense to alot of people. You can’t globalize products, shift hours to part-time status, reduce wages, cut benefits, while monetizing greater incentives (with less punishment) at the wealthy end without the data at the point of release starting to not make sense. The only difference is that the damage from an earthquake is fairly time-constrained; after awhile, you know the shaking has stopped. But with the structural changes taking place and even accelerating, this kind of shaking isn’t going to end anytime soon. And that means, you have to be even more circumspect about any data that says otherwise. You can’t deal with the present if you keep looking for indicators that point to the past.

  2. Kathryn Baer says:

    Isn’t it the case that the U-6 undercounts discouraged workers? I’ve read that they’re excluded after a year of being so classified.

    • Jared Bernstein says:

      Right–to be “discouraged,” you have to have looked for work at some point within the past 12 months.

  3. Larry Signor says:

    The noise of how to measure employment/unemployment seems to have overcome the fact that we don’t have enough jobs, no matter which metric is used. Most discouraged workers are waiting for the jobs to come back, which has been our historical experience. This may not happen this time and it is very confusing to older ex-blue collar workers. We didn’t leave the job, the job left us.

    • Robert Buttons says:

      In a global economy, a large slate of “workers rights” leaves you with a large slate of rights but no job. Spain is a perfect example.

      • Robert Weiler says:

        Or Germany or Austria. Oh wait, they have a large slate of workers rights and jobs. Perhaps it isn’t as simple as you would like it to be.

        • Robert Buttons says:

          Interesting you should cite Germany, a country with no minimum wage. Let me just say workers’ rights grow out of prosperity, but prosperity does not come from workers’ rights.

  4. Robert Buttons says:

    There’s fixed definition of “unemployed”. There’s no fixed definition of “GDP”. There’s no fixed definition of “inflation”. There’s no fixed definition of “money supply”.

    But when we perform mathematical functions on these fuzzy numbers, we are told real answers result (bonus pts if calculus is used). That’s called econometrics. pffffffftttt!

    • jonas says:

      Why do you say that “There is no fixed definition of…?” Is there something I am missing?

      • Robert Buttons says:

        The measurement of these economic variables is rather arbitrary and is completely uscientific, when compared to, say, measuring the acceleration of gravity.

        • jonas says:

          Unemployment: “Persons are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work.” via:

          GDP: “GDP is the market value of the final goods and services produced by labor and property in the United States. GDP is equal to the sum of personal consumption expenditures, gross private domestic investment, net exports of goods and services, and government consumption expenditures and gross investment. GDP is also equal to the sum of value added by industry across all industries.” via:


          Money Supply:

          It’s one thing to not understand these terms; it’s another say that these measures have no ‘fixed definition.’ But so long as you know better than the BLS, the BEA, and the Fed, I guess we should believe you.

          • Robert Buttons says:

            1. The links you provide have DESCRIPTIONS, not DEFINITIONS. For example a description of a meter would be: A metric unit of measurement. A definition would be: the distance traveled by light in 1/299,792,458th of a second. Please define GDP and CPI.
            2. Since GDP is typically given in units of dollars, define dollar.
            3. If “unemployment” is a fixed definition why are there U3, U6, etc?
            4. Here’s the link description of “money supply” “M2 is defined as M1 plus savings deposits, small-denomination time deposits (those issued in amounts of less than $100,000)”. Please explain, (without being arbitrary!!!) why deposits of $99,999.99 count towards M2 and those of $100,000.00 do not?

            These should be easy questions to you, as you clearly have an understanding. Or am I missing something?

  5. Dave says:

    There are a lot of cynical people who believe that people giving up and dropping out is a good thing, because they are the least desirable workers. Of course, it is absolute nonsense that those that have dropped out are the least desirable. They’re for the most part just unlucky.

    There are a fair number of smart people that think the economy just has to shrink for some reason, and so they don’t think this is a big deal. I think this is the biggest psychological barrier to getting the message across that this is a really, really bad thing.

  6. urban legend says:

    U-3 may even be behind the times as to what constitutes a job search. Says BLS: “Passive methods of job search do not have the potential to result in a job offer and therefore do not qualify as active job search methods. Examples of passive methods include attending a job training program or course, or merely reading about job openings that are posted in newspapers or on the Internet.”

    Various on-line services — Craigslist, Indeed, for example — are a zillion miles more thorough than newspapers and other services once were. Consulting them is the best way to find actual openings. But when there aren’t any jobs remotely in your wheelhouse showing up in these databases, there’s no point in doing other things just to spin your wheels other than to receive unemployment. There’s only so long you can send out 500 resumes and get nothing more than a handful of “keep your resume in our files” responses. There’s only so many times you can hit up your contacts and ask for new people to talk to, too.

    If you out of a job but are continuing to check the good databases on a fairly regular basis, that should be considered active searching. It certainly means you are unemployed in any meaningful sense.

  7. jonas says:

    The chart in the linked blog post (sorry; don’t know if I can or how I would post an image here) shows the gap between the number of unemployed and the number of job openings. We’re still north of 7 million on this measure, and “UNEMPLOY” does not count those counted in U6.

    Alternately, there is the Employer Market Power Ratio. ( This measures the number of people (theoretically) competing for each job. Again, we are not counting discouraged workers here.

  8. Noni Mausa says:

    Michael said: “…The aggregate erosion of a middle-class, factory-based, labor force, like mini-earthquakes hiding the major earthquake building up behind it, is going to give data crunchers alot of semi-false data. Eventually, like a major earthquake, the sudden change in the data isn’t going to make sense to alot of people…”

    The mini-quakes have been happening for the past thirty years, but early on they were cushioned and concealed by the resiliency of a healthy middle class. Just as a person can skip a meal with no effect, the early adaptations were minor, easily perceived as individual accommodations, not symptoms of a widespread change. The camel doesn’t notice the first straw, nor the hundredth… but eventually (unlike boiled frogs, who are much stupider than camels) he will lie down and quit. This looks like a sudden change (working camel => reclining and/or dead camel) but could have been forseen long before the camel became terminally distressed.

    I wonder if avalanche math could be applied to economic topics?

    Predicting the upcoming big avalanche is a mugs game, but here are my guesses as to how this generation will collapse to a new lower energy dead-camel state:

    — as the last of the boomer generations die or go into care, the cushioning they have provided for their children and grandchildren will vanish
    — traditional, fixed benefit pensions of the older generations, their home equity, and whatever resources they retained after the 2008 housing and investment collapses, will vanish, if they have not done so already in helping out the kids
    — the vast resource pool of parents’ basements will not, presumably, vanish, however they will not be the “free” dwellings they have been.
    — house prices will take a dive, as buyers lose the ability to buy or maintain a single family dwelling
    — as the unsupported, resource-poor postwar generations continue to lose traction, we will see a flip into boarding houses as entry level dwellings. Workers will be poorer, more mobile, less tied to a specific town or area. Will we see chains of corporate boarding houses, providing housing much like college dorms?

    Hmm. And now I wonder, what will happen to congressional districts, if workers become fluidly nomadic?

    Anyway, those are some of my guesses. I am not guessing widespread riots and rebellion, for at least another 25-40 years.