They aren’t coming back
Back in 2013 I argued that the low Labor Force Participation Rate was not evidence of lots of labor market slack:
It’s true the payroll gains and falling unemployment rate overlook the low labor force participation. In my view the falling LFPR is not a cyclical issue, even though the variable is itself cyclical. This is very confusing to most people. Imagine a LFPR that has a strong downward trend for structural (non cyclical) reasons. Also assume the actual LFPR falls in a more cyclical pattern, falling steeply during recessions and leveling off during booms. The level periods look like “nothing happening,” whereas the LFPR is actually growing relative to the declining trend line. My prediction is that the LFPR will stay low even after we recover from the recession, and we always recover from recessions. It’s not a cyclical problem. This will become obvious by 2016.
Last October I explained the point further:
I noticed that the labor force participation ratio fell to 62.7%, the lowest rate since February 1978. Folks, it’s not coming back. In less than a year the recession will completely end and we will get a normal unemployment rate (about 5%). Jobs will be available and those people simply aren’t coming back. They are early boomer retirements (perhaps discouraged by the previous job market), disabled (perhaps partly discouraged by the job market in previous years) and young people staying in school longer, or choosing to work less (as is true in affluent towns like my own Newton, Massachusetts.) It pains me to say this but it’s pretty clear they aren’t coming back””the job market is good enough where the LFPR rate should not still be falling, if it really were nothing more than discouraged workers sitting there ready to plunge in again when things got a bit better.
Now it’s 2015 and the progressive mainstream media is finally beginning to contemplate the unthinkable. Here’s the (Keynesian) New Yorker:
Despite the subsequent economic recovery, which has now lasted for more than five years, the rate has continued to fall. Last month, it stood at just 62.7 per cent, a tie for the lowest level since 1978 (a time when more women stayed at home and did domestic labor rather than join the official workforce). . . .
Most policy-makers, including Janet Yellen, the chair of the Federal Reserve, have been assuming that much of the decline is cyclical and that, as the recovery picks up, more and more discouraged workers will return to the labor force. . . .
I agree with that argument: indeed, I’ve used it myself. By now, though, we should be seeing signs of the participation rate rebounding. The fact that it isn’t is somewhat alarming.
But not surprising to readers of TheMoneyIllusion.
Tags:
11. January 2015 at 08:33
Well not sure about this one. Societies are and businesses adapt, are mutable. People can choose to retire later or work as consultants if the demand for labor is there.
If wages rise, should not more labor be supplied?
No slack in the labor force? Then why are unit labor costs falling?
I wish we could conduct a natural experiment and have such an aggressive growth-oriented Fed that we experience “labor shortages” — then what would be the labor participation rate?
11. January 2015 at 08:44
Jobs exist to serve some purpose, not to employ someone. There are lots of Boomer jobs that Millennials just don’t want to do, because their living standards have always been so high. That’s why we get WSJ stories about $140K welders and people Mike Rowe exist, even as we import huge numbers of foreigners for IT jobs (not complaining, I married one).
When you consider that Americans consuming $30K are global one-percenters, the employment picture suddenly makes perfect sense.
11. January 2015 at 08:59
Ben, I actually think there is slack in the labor force, but I don’t think the LFPR is the right measure of it. I’d look at the U3 unemployment rate, and maybe the U6 rate too.
11. January 2015 at 09:31
More Sumnerian double-speak. Like the Roman god Janus, Dr. Sumner speaks with two faces: “Imagine a LFPR that has a strong downward trend for structural (non cyclical) reasons. Also assume the actual LFPR falls in a more cyclical pattern, falling steeply during recessions and leveling off during booms. ” “I agree with that argument: indeed, I’ve used it myself. ” “Ben, I actually think there is slack in the labor force, but I don’t think the LFPR is the right measure of it. ”
Or, instead of the above complicated gymnastics and dualities, you can assume the LFPR is simply depressed for the same reason money velocity is depressed and the same reason inflation is depressed: good ole old-fashioned lack of demand…
11. January 2015 at 10:18
I think the best measures of the labor market situation are
research.stlouisfed.org/fred2/graph/?g=EUc
and
research.stlouisfed.org/fred2/series/LREM64MAUSM156S
My questions are: what made the recoveries from the 2001 and 2008 recessions so weak? And why is the situation even worse in the U.K.? This article has part of the answer, but it can’t be all of it:
nytimes.com/2012/04/13/us/13iht-letter13.html
I doubt the problem is demand-side; prices should have adjusted by now.
11. January 2015 at 10:26
Note: the situation in the U.K. was better in terms of the labor market, but much worse in terms of RGDP growth and real wages.
11. January 2015 at 12:54
I’ve done a spreadsheet projection based on age-specific participation rates leading up to the Great Recession, when participation rates were steady (except for a slow rise among age 60+ workers). That way I can eliminate changes in the population structure, particularly baby boomer retirement.
The data I have (from the CPS) let me do breakdowns by 5-year age brackets. For brackets from age 25-29 to those age 50-54 the movement is quite similar, from 100% in early 2007, falling after May 2008 and plummeting from October to an average of 94% of “normal” in February 2010. While not smooth, participation began rising from January 2013. In December 2014 it was up to 97%.
So true, the baby boomers aren’t coming back. And participation for new school leavers (age 20-24) is still only 91% of normal, up from 86%. But the data do suggest that we are in fact slowly returning to normal if we’re careful to adjust for the shifting structure of the US population.
Now 3% of “normal” is still millions of people. (I only have percentages in my spreadsheet so can’t make that more precise.) In the end we will look like Japan, having lost a decade….
11. January 2015 at 13:12
[…] Scott Sumner goes back once again to the Labor Force Participation Rate: […]
11. January 2015 at 13:20
If these people aren’t coming back, that would mean there isn’t slack in the economy. That would in turn mean we have tight labor markets and we would see wage inflation. But we don’t. Supply and demand.
I don’t buy these demographic reasons and theories tossed about without evidence. It’s an awfully big coincidence that they all hit right after a huge cyclical event.
11. January 2015 at 13:53
Ray, I always love it when commenters attribute statements to me that were made by others. Especially using quotation marks.
Thanks Mike, That seems plausible.
Peter, Reread my first quote.
And there is slack, it’s just that LFPR is not the right measure of slack, use U3 or U6.
11. January 2015 at 14:13
One thing I will posit is some the LFP change is due to substantial changes in most corporation’s HR policies. There is considerable evidence that jobs remain open far longer than before and corporations interview many more people per position than before.
The big change starting in 2000 was companies had far more resumes coming in and sifting through. I wasn’t around for the pre-Internet job era, but the real work required limited the applications. Now HR departments are subbed out to go through all the applications. That new model particularly hurts less-than-ideal candidates, such as the currently unemployed or those trying to change careers.
It’s not a theory completely thought out with economic theory. But if the frictions are stong enough for discouraged workers, then higher NGDP could lead to higher inflation despite a ready supply of discouraged workers willing and probably able to work. The companies which do take a chance on the discouraged workers could charge lower prices than the companies that don’t, but what if no company takes that chance? Most industries have economies of scale which prohibit discouraged workers from starting new businesses and if no companies take a chance on discouraged workers, then discouraged workers will remain discouraged.
11. January 2015 at 14:14
Mike, that’s similar to what I have found. I agree that there is still some slack. Lfp will likely level out for a couple of years until it converges with the downward sloping trend.
But, on caveat I would suggest to your analysis is that labor markets were very hot in 2007. This was masked then by the downward trend in lfp, just as it is now, so that it isn’t widely recognized, even by those who claim that the 2007 economy was in a bubble in other contexts. So, there was probably some opportunistic cyclical employment at the point where you are pegging your trend line to 100%. So, we are probably looking at a recovered normal lfp even if your measure recovers to 98 or 99%. My 2 cents, anyway.
11. January 2015 at 16:07
If we think of the LFPR as a supply curve, then we will get higher LFPRs at higher demand levels (as expressed by wages).
There is some contention on this, but wages have been stagnant in the US for 40 years. Labor share of business income has declined from 70% to 60%.
FICA and VA “disability” programs have exploded.
Actually, maybe we should be surprised anyone is working….
11. January 2015 at 22:44
So then, explain cross country comparisons for me.
As we see on my post (below), the US is the only advanced country to suffer employment-to-population declines of the magnitude we have seen here. It is not bad, the US is absolutely the worst. Worse than Italy, the UK, or France.
How is it that the Germans were able to raise their emp-to-pop ratio even as the US fell, when the Germans are aging out? How is it that the emp-to-pop ratio is 7 pp higher in the Netherlands and 12 pp higher in Switzerland? I thought these countries were all aging out, and yet they are trouncing us.
Do explain.
http://www.prienga.com/blog/2014/10/31/employment-to-population-ratio-detail
11. January 2015 at 23:12
Did the helicopter drop $100 billion last year?
http://www.wsj.com/articles/the-fed-cash-machine-1421018574
12. January 2015 at 00:47
Steven,
It’s because 50-60 year olds have lower LFP in the US, partly because these are prime years for disability eligibility. So, as baby boomers move through this age group, the demographic decline in LFP is sharper in the US than it is elsewhere.
12. January 2015 at 00:47
Scott, have you seen this?
http://www.voxeu.org/article/household-credit-and-employment-great-recession
12. January 2015 at 06:12
Ben, Higher wages might lead to higher LFPR, but won’t monetary stimulus boost wages and prices, leaving real wages little changed?
Steven, I’m just telling you what is going to happen (LFPR won’t recover) I’m not telling you why. In the case of Germany, labor market reforms boosted LFPR—we could do the same. The US labor market has lots of structural problems. I haven’t studied the differing demographics, and hence don’t know what role they play. LFPR is not coming back—but I have an open mind as to why. You tell me.
12. January 2015 at 06:15
Vivian, And the Fed was afraid to do QE because they feared losing money on the program.
Kevin, I’d just point out that explaining variation across counties in employment losses is a very different proposition from explaining overall employment losses (which depend on NGDP and nominal wages.)
12. January 2015 at 07:25
As for LFP:
It’s principally driven by policy; it’s absolutely clear from the graph. Changes in the LFP rates are dramatically different in the US than any other OECD country–even as the US economy has out-performed pretty much all of them.
The factors in low and declining US LFP will be:
– student loans (unwinding now)
– disability, as Kevin points out, still getting worse
– welfare (seems to have become more liberal)
– taxes on spouses (not clear on this, but may be a factor in fewer women working)
So it’s mostly related to paying people not to work. LFP is an endogenous variable. You can make it whatever you want, certainly up to the 70% threshold. In public policy terms, the current US LFP rate should be considered scandalous, which it is.
You know, it’s not like I don’t have a spreadsheet on this. And it’s not like I haven’t linked it, what, three times before. Mr. Erdmann could make a nice living for his blog for a week just running some numbers. And Scott, you could do worse than cracking it open and putting together a little model which integrates population growth, legal and illegal immigration, civil labor force, labor force participation, and disability, et al. (See the “Analysis” tab.) It will give you a very good feel for how all these numbers move together, and you also have all the data to look at age cohorts and the impact of education. I have done all the hard work for you.
In any event, if I take your view, Scott, then the US economy will be gated by a shortage of labor within maybe 12 months. And then I have no idea why we’re interested in NGDP targeting, because it will be entirely beside the point if we’re at full employment. It will be Secular Stagnation 2.0 per Larry Summers (he disavowed 1.0), that the US will run out of labor before it reaches…ummm….potential GDP.
I think Larry will have an opportunity to disavow 2.0, too.
Here is the spreadsheet again: http://www.prienga.com/blog/2014/10/3/employment-tables
12. January 2015 at 08:08
Scott,
I have only seen the summary. But, my understanding of the paper is that the situation with Wachovia created an exogenous shock to credit. He finds that there are frictions in credit markets, housing markets, and economic activity in general that created local effects proportional to Wachovia’s local market share.
In effect, each local area had an independent monetary policy, resulting from this. I see this paper as evidence of the monetary nature of the crisis.
12. January 2015 at 13:59
Steven et al
Only a really wealthy country like the US can pay so many people fit to work not to work.
🙂
GDP per capita should fall, or grow relatively more slowly than “tougher” countries more as a result. Unless offset by stronger than average productivity growth by those in work in the US. Which is also possible while it remains the global home of tech advances.
12. January 2015 at 14:59
Steven Kopits’ reply gives far too much attention to public policy, and not nearly enough attention to the simple reality that the US is a wealthy country, and in a wealthy country, wealth and income is more likely to leak out from rich to nonrich and working to nonworking people in a variety of methods, including private transfers.
There’s a lot more real suffering and a higher likelihood of a lower standard of living without employment in america, relative to pakistan. But you couldn’t fix that without literally turning america into pakistan. Most people would say that that isn’t worth it. Public-sector behavior is only a small piece.
I also very much doubt Steven’s general point: LFP is on the decline in every advanced country, or will be, in the long-enough run, because our labor is too expensive relative to alternatives, because living in america is expensive, because we’re a wealthy country. The end. Rising productivity, more goods per person, but finite and downsloping marginal interest in buying those goods per person and/or asset accumulation per person. Falling demand.
Can’t be fixed, shouldn’t be fixed, won’t be fixed.
12. January 2015 at 15:18
Scott Sumner: “Steven, I’m just telling you what is going to happen (LFPR won’t recover) I’m not telling you why.”
Kevin Erdmann: “It’s because 50-60 year olds have lower LFP in the US”
Why is no one mentioning the effect that ObamaCare is having on the 50-65 set, without triple digit effective marginal tax rates and huge subsidy cliffs? And yes, these tax rates apply to spousal income, too.
12. January 2015 at 15:53
Steve, while I agree with your sentiment, I don’t think the trends and timing of trend changes match that explanation.
13. January 2015 at 07:27
Glasnost –
Dead wrong. If your hypothesis were right, then LFP rates should be lower in Germany and Switzerland than in the US. They are much higher, and Germany’s has increased as much as the rate has fallen in the US. The data flatly refutes your assertion.
Kevin –
Look at the graph. Read the graph. Interpret the graph. Nobody better at that than you. Tell me what the graph says, Kevin. I think it allows only one interpretation. But if you have two, let’s hear it.
http://www.prienga.com/blog/2014/10/31/employment-to-population-ratio-detail
Scott –
Well, I don’t know what to tell you. In your world, this economy is done by year end. Is that what you believe? Do you also subscribe to Summers Sec Stagnation 2.0? If not, how do you overcome a structural shortage of labor? How do you accomplish that without substantial wages increases?
In any event, if I accept your view, maybe Ken Duda should sponsor a center to encourage people to go back to work. (That’s my wife’s business, taking her turn on NPR this morning. http://www.wnyc.org/story/casino-worker-struggles-steer-her-family-past-atlantic-city-crisis/ She’d take Ken’s money.) It’s not as though we lack a template for these labor market reforms. You can see them very plainly on the graph.
13. January 2015 at 07:44
Steven, You said:
“In any event, if I take your view, Scott, then the US economy will be gated by a shortage of labor within maybe 12 months. And then I have no idea why we’re interested in NGDP targeting, because it will be entirely beside the point if we’re at full employment.”
The purpose of NGDP targeting is not to get us to “full employment”, it’s to smooth out the business cycle and produce greater nominal stability.
You said:
“So it’s mostly related to paying people not to work.”
I find that plausible, although it’s also aging boomers like me.
How do you explain the fact that per capita GDP in places like France is only 70% of US levels, and falling ever further behind?
Kevin, OK, But I’d prefer to call that “credit policy.”
James, Good point.
Glasnost, Wealth is a part of the story, but far from the whole story. The LFPR was rising in the 1990s, a prosperous decade.
Steve, I’ll mention it–Obamacare is a (modest) part of the story.
Steven, I don’t agree with Summers, I think AS is the problem, not AD. And no, I think output will keep growing, but growth will slow after 2015.
13. January 2015 at 08:34
Steven that is an interesting graph. But it’s the other countries that surprise me. Who would have thought that Europe is generally seeing a rise in employment?
13. January 2015 at 08:44
Yes, I agree on NGDP and stability. It’s another tool, and perhaps helpful.
Notwithstanding, the labor issue matters, Scott. It matters a lot if it’s the binding constraint.
Here’s what I see as I have been doing the partial forecasts: a boom in blue collar employment not seen since the 1960s and real strength in demand over several years. Just for example, a normal airline recovery would require something more than an additional 1,500 commercial aircraft over three years just to cover growth in the US. That’s not normal fleet turnover, just catch-up growth. That’s as much as the combined annual output of Airbus and Boeing, worldwide. And that’s not including normal US fleet turnover, China growth, European recovery–none of it. We’ll see the same forces at work in other markets, too.
And there are airports and asphalting (+40% next three years), heavy truck construction (+25% next two years), and by the way, increased shale oil production (+20% annually from 2016)–which we’ll need about a year from now. We’re almost a decade behind on infrastructure, and that’s going to come back and bite us–in a good way. There’s going to be a lot of work if oil prices stay low, and a lot of it will be old fashioned manual labor.
These gaps are really impressive. So, if we don’t have enough labor, it’s going to be an issue. A big issue. And that has ramifications for LFP, legal and illegal immigration, older person LFP, etc. I would not underestimate just how big a boom may be headed our way if oil prices stay low.
Let me close with a couple of factoids:
Low oil prices will increase US GDP by 0.5% (pp) in 2015. Just the primary price effect, and just on net oil imports.
For Vietnam, that number is 3%; for India, 2.4%; and for China, 1%. That’s a hell of a lot of GDP we’re talking about.
13. January 2015 at 09:56
Steven: Do you think we have the labor pool for those projects?
I don’t see that unemployed 50-year-old men, women, white collar service workers, etc. are going to be able to jump into service as welders or even ordinary laborers.
Even young people today aren’t growing up working on cars, working with their hands, etc., and might not be have the skills to do those jobs needed.
13. January 2015 at 10:37
Well that’s the question, Charlie. That’s why I’m pushing Scott, because I think we need to stop and reconsider here.
So what happens if we don’t have enough labor? Well, wages rise. Maybe by a lot. Is that happening in isolation, or are inflation and interest rates picking up as well?
If we don’t have local labor, what happens? Well one thing is that we get a huge wave of illegal immigration. (See the spreadsheet!) If so, you better be taking a hard look at your immigration policy, because they’re coming whether you like it or not.
And labor force policy is part of that. If you’re paying people to sit on their hands, then you’ll get more illegal immigration.
And what about infrastructure? It gets incredibly tight. So maybe we should borrow $300 bn in long-term govt bonds while interest rates remain low and make that available for infrastructure expansion over the next five years.
There’s a lot going on potentially. We’ve become used to thinking of the US as a spent nation. The truth might be just the opposite. We are coming out of a near depression, and we’re going to see the catch-up that we’d expect we such a rebound. That’s also something to think about.
13. January 2015 at 18:54
Maybe we should cut the payroll tax and increase the EITC to lure some of those retirees and disabled into the labor force.
13. January 2015 at 19:16
Vest social security if people don’t retire.
14. January 2015 at 07:17
Steven, One factor that I probably have not paid enough attention to is immigration. A boom in the US might suck in a lot of immigrants from places like Mexico.
14. January 2015 at 12:22
It’s in the spreadsheet, Scott.
Illegal immigration runs about 500,000 in down years, perhaps 1.5 million in an up economy.