Skip to content

Raw data: Prime age workers in the post-COVID era

How's the ol' labor force doing? The Wall Street Journal says today that "several million" workers have dropped out of the workforce permanently thanks to COVID-19, but I suspect that yet again this is a misconception that comes from looking at the entire labor force instead of just the prime-age labor force. Here are the prime age labor force levels and participation rates for both men and women:

Among men, both the labor force level and the participation rate are right on trend for prime-age workers. Among women, the labor force level is below trend but the participation rate is slightly above trend. Here's how that pencils out:

Among women, there are 0.3 million fewer prime age workers than the trendline projects. Among men there are 0.2 million fewer.

In other words, nothing. These differences are so small they're within the margin of error of the BLS survey. Right now the prime age labor force is exactly at its pre-pandemic trend level.

Now, there are about 1.6 million prime-age women who have exited the labor force (top chart), though they were probably the ones who were most weakly attached in the first place and will only be coaxed back by higher wages.¹ And there might be more still outside of prime age. So the Journal might be correct—but about a fairly meaningless metric. The one that really matters is prime-age workers, and on that score we're pretty much at full employment.

¹Which isn't happening no matter how many McDonald's signs you see that are allegedly offering $20 an hour to flip burgers. Despite all the complaining from employers about how hard it is to find workers, real wages in most industries have been falling for the past year. Even in the restaurant biz, which is supposed to be the tightest around, real wages have only increased about 6% over the past year. In other words, burger flipping has gone from around $12 an hour to $12.70 an hour. That's hardly going to tempt a lot of marginally attached workers back into a fast-food kitchen.

25 thoughts on “Raw data: Prime age workers in the post-COVID era

  1. D_Ohrk_E1

    I have may have been watching too many Peter Zeihan videos. What attracts me to his videos is that he's pulling lots of disparate data points together, which is not something that is easy to do, but if you want to predict a given "butterfly effect", you need to incorporate a ton of data points, not just a handful.

    This video portion -- https://youtu.be/F2C42p_8XfI?t=405 -- is a theme he repeats quite often re baby boomers and their effect at retirement and immigration changes. I don't accept all of his ideas, but I generally accept the wider trends he's describing, specifically re baby boomers. It is why I strongly believe that the threat of the zero lower bound (ZLB) will come back with a vengeance as soon as we recover from this exogenous shock.

    Looking at the prime age workforce participation rate is not very useful. Compare the workforce participation rate by generation, instead. You've done this before. Tell us that story instead.

      1. Vog46

        D_Ohrk
        No apologies needed

        Here's another perspective from the St Louis Fed
        https://www.investopedia.com/terms/p/participationrate.asp

        There is a wonderful chart in there
        From the associated article:
        {snip}
        Demographic factors
        Changes in the working-age population from generation to generation influence labor force participation as well. As large age cohorts enter retirement age, the labor force participation rate can fall.

        The retirement of a steady stream of baby boomers has reduced labor force participation. According to the Federal Reserve, the share of prime-working-age people (25 to 54 years old) in the labor force peaked at 72% in 1995 and declined to 63.7% over the next 25 years.67 This roughly corresponds to some of the declining trends in labor force participation in the 21st century.

        An increase in college attendance at the younger end of the age spectrum is another factor that reduces labor force participation. College enrollment by 18- to 24-year-olds increased from around 35% to 41% from 2000 to 2018.8 However, enrollment rates have dropped due to the pandemic, with undergraduate enrollment declining 7.8% from fall 2019 to fall 2021.9

        Global Labor Force Participation
        Global labor force participation has shown a steady decline since 1990. According to the World Bank, the global labor force participation rate stood at 58.6% at the end of 2020, down from 62.4% in 2010.10
        {snip}

        COVID took MILLIONS of people outside of "prime" aged workers out of the work force. They were on the verge of retiring but COVID propelled them to retire.
        AT THE SAME TIME those aged 16 to 24 have a MUCH higher % of their numbers in college. Good for us in the long run but bad for statistics that have been based for the last 50 years on a baby boomer fed population number.

        We are so used to seeing these numbers we assume they've been around for forever but the Dept of Labor has only been around since the 1970s. Other entities have kept track of unemployment longer than that but the Bureau of Labor stats don;t have that longevity.

  2. Zephyr

    I don't understand this argument. Why focus only on prime age workers? If a business loses workers, no matter what their ages, they've lost workers.

    1. Jasper_in_Boston

      I was wondering the same thing. Focusing on prime age workers makes sense for some purposes. But I wasn't quite sure why Kevin focuses on them for this particular point.

  3. Justin

    What are the implications for economic policy? Inflation is causing the federal reserve to increase interest rates. They hope to make housing even more unaffordable and get a bunch of people fired so there isn't so much demand for fast food chicken sandwiches. If we are at full employment, then we can't afford it. Decades of campaign ads touting this or that policy to "create good paying jobs" are now fulfilled and everyone hates it! Now what?

    1. Austin

      Don’t feed this troll. In other posts he calls for more attacks on Russia, and is cool/indifferent with Russia responding by nuking a few US cities… so I’m sure he’s arguing in good faith now about his “worries” about the economy.

    2. skeptonomist

      There have been many, many years since the 80's when unemployment has been below the supposed level that triggers inflation (the NAIRU) and it hasn't happened. "Full employment" has not caused inflation. Other factors are causing it now. The supposed correlation between inflation and unemployment, the Phillips curve, blew up in the 70's and 80's - we had both high unemployment and high inflation. These things just don't work the way economists assume they do - they ignore the evidence which contradicts their assumptions.

      1. KenSchulz

        I’m shocked - shocked! - to hear that economists have more fondness for their theories than they have for data!

  4. Austin

    Like Zephyr says, unless the non prime age workers were doing nothing essential, wouldn’t a few million of them vanishing from their jobs make a difference in how those businesses function?

    For example, elderly greeters at Walmart seem to have completely vanished yet Walmart carries on just fine without them - but certainly not all workers over 54 were doing tasks that turned out to be “tasks” easily eliminated without being missed. If a business/industry had a disproportionate number of non prime age workers, wouldn’t it stand to reason that it might be struggling more now with the disproportionate disappearance of that age group from the labor pool, even if all other businesses/industries are fine?

    1. Zephyr

      My local supermarket employed a bunch of older-than-prime-age workers, and now they seem to be compensating by changing a bunch of check out stations to self-service only. The average age where I live is close to 54 so there are a lot of non-prime-age workers.

    1. Austin

      Kevin has written about prime age workers numerous times before, always using the Bureau of Labor Statistics definition of ages 25-54. Perhaps you’re new here and to the world of labor stats, but Kevin usually does reiterate somewhere in his posts the ages considered prime. This time, with his previous post about getting knocked on his ass by a new treatment drug, he appears to have neglected to do this.

    2. Austin

      If you Google “prime age workers US” you’ll see that virtually every website agrees that prime age = 25-54. This definition appears to be so common that Kevin probably just assumed everyone knew it already.

      Similar to how few news articles nowadays stop to explain that “seniors” or “senior citizens” are “people over 65.” I’m sure there are some people in the US and lots more abroad who don’t know what constitutes a senior citizen, but it’s rare for journalists to explicitly spell out who is in that group.

  5. rharrisonauthor

    My wife and I a great example of "non-prime" workers who said "sod-it" because of covid. The callousness of our university, where we were both tenured and funded full-professors, with respect to health risks made it clear that we weren't exactly considered valuable employees. Unlike I may add, the administrators who still work from home. So we just accelerated becoming emeritus. I suspect more than a few people in our position, able but not ready to retire, simply decided that enough was enough.

  6. skeptonomist

    I don't see why the "prime-age" fraction, presumably 25-54, is so important - there are still lots of people outside these limits, But here is a measure using the 15-64 limits:

    https://fred.stlouisfed.org/graph/?g=Ok8v

    you may need to hit "max" to get all of these graphs
    This shows less than a million missing from the pre-covid level so there can't be "several million" who have permanently dropped out. We can also look directly at ages 55-64, the (large) group which might have retired prematurely:

    https://fred.stlouisfed.org/graph/?g=Ok8w

    Again nothing, though there is a strange jump in January.
    Measures of employment are tricky - there are several things that may come come into play. It looks like the WSJ made a big mistake but any conclusions on this may have flaws.

    1. skeptonomist

      These are actual employment numbers, not "labor force", but unemployment level is similar to pre-covid so the results are similar to Kevin's.

    2. Vog46

      What is hard is that if we wish to take a look at 55-64 then we also have to compare pre and post pandemic numbers for those 65+ who also left the workforce
      Then you have the problem of the questions asked on the surveys. If a person 62+ leaves the work force and is NO LONGER looking for work they are dropped from the rolls - even though the POSITION they occupied may be open.
      Employment surveys are hard
      But keep this in mind
      There are fewer "prime age" workers than there were when BLS started collecting data. This number of prime age workers will CONTINUE to drop as the population growth tapers off (we are currently LOSING more people than are being born). If technology cannot improve enough to take over many of those labor tasks something has got to give.
      This puts the job creation efforts of Clinton and Reagan in a new light. They created enough jobs to absorb the boomer generation. Obama's job growth is slightly lower (which is a good thing) but we need to reset expectations to reflect a smaller and smaller work force
      I won't be around for it but I believe the effects will be profound.

  7. Spadesofgrey

    Be careful with "wage growth" vs inflation hedonics. They are revised. 1 978-79 is a great example of busting high on inflation and giving bad data. Which led to bad policy.

  8. akapneogy

    "So the Journal might be correct—but about a fairly meaningless metric."

    Isn't the Journal's mission to be correct about meaningless metrics and incorrect about meaningful metrics?

  9. Pingback: Interesting labor statistics | Later On

  10. spatrick

    The fact that there's even a discussion on "prime age workers" basically points out what's been true about the U.S. economy since the 1990s: that its function largely depended upon a massive, cheap pool of labor, not just immigrants but bored retirees looking to pick up some extra income that their Social Security, 401Ks, pensions and anything else couldn't provide for them. COVID-19 hits and that's all gone. Thus fewer people in the work force, more need for higher wages to attract workers for jobs that normally went filled by the old and or immigrant and thus higher inflation.

    Bottom line is you want to restrict immigration, be prepared to pay the extra costs for it.

  11. Pingback: Raw data: The child day care worker shortage – Kevin Drum

Comments are closed.