A missing explanation for why the National Journal says we have missing jobs
Tyler Cowen and Arnold Kling both linked to a National Journal article that implied America had lost (or failed to create) 15 million jobs in the last decade, even apart from the effects of recession. Go back to April 2008, when unemployment was a quite low 4.9%. The claim is that even then we were suffering from 10 to 15 million missing jobs. The data is a bit vague, so it’s not clear exactly how many:
The Great Recession wiped out what amounts to every U.S. job created in the 21st century. But even if the recession had never happened, if the economy had simply treaded water, the United States would have entered 2010 with 15 million fewer jobs than economists say it should have. . . .
The forecasters said [in 2000] that the economy would create 22 million jobs over the next 10 years. At the decade’s economic peak, though, that number stood at only 7 million. Job growth in the 2000s was the lowest of any decade ever recorded by the federal government, stretching back to the 1940s. As a result, workers were extremely vulnerable to the tidal-wave recession that washed away all of the decade’s meager gains.
How could that be? As a matter of pure arithmetic, missing jobs imply one of two things, a huge rise in unemployment, or a huge fall in the labor force participation rate. Can you think of a third factor?
Since unemployment was only 4.9% in April 2008 (i.e. normal) there should have been a huge drop in the LFPR between 2000 (which was a red-hot boom year), and 2008. Yet according to this graph, it looks like the LFPR merely edged down from about 67% to 66% between 2000 and 2008. I’m surprised the decline was that small, given how the dot-com boom sucked anyone who could tie their shoelaces into the labor force. Remember what fast food help was like back in 2000?
I notice that the LFPR rose sharply between the 1960s and 1990, I’d guess due to more women working. Then it leveled off at about 66.5% during 1990-96. Then it rose to 67% during the dot-com boom of 1998-2001. Then it dropped to 66%, and leveled off until 2008. Aren’t there lots of possible explanations for this tiny drop in the LFPR? Recently I’ve noticed more adult women who could be working, but choose to stay home. I don’t know if that’s a trend, but Jim Tankersley doesn’t provide us with any of the data we’d need to make sense of the claim about missing jobs. He may be completely correct; I just can’t see where the numbers come from.
Now if you go up to 2010, then yes, I do see a worrisome loss of jobs. This shows up as both much higher unemployment (perhaps 8 million lost jobs), and a bigger drop in LFPR (another 2 million lost jobs), both obviously related to the late 2008 plunge in GDP. I just don’t see evidence that we had a massive jobs problem before the recession. Does anyone know what data can support Tankersley’s claim, and why it doesn’t show up in either the LFPR, or the unemployment rate in April 2008?
I still think our unemployment problem is about 80% AD and 20% structural problems (99 week UI extension, 40% minimum wage jump, etc) but I have an open mind on the proportions.
PS. In my earlier housing post, I should have cited this Nick Rowe post on a housing “Phillips Curve.”
Tags: Unemployment
21. January 2011 at 11:58
“Can you think of a third factor?” Emigration?
21. January 2011 at 12:28
Yeah Scott, please address immigration especially this Reuters story/post:
http://www.reuters.com/article/idUSTRE70J37P20110120
21. January 2011 at 12:37
The figure you added tells the story you find missing. In the decades from 1960s to 1990s the labor force participation inched upwards – there’s your smoking gun. The National Journal article merely assumes that the labor force participation would go up as much as it did in previous decades. Hence, the jobs that were never added.
Draw a trendline from 60s to 90s and continue it to 00s and you’ll find the answer to your mystery.
That and then possibly immigration.
21. January 2011 at 13:06
Scott,
You wrote:
“How could that be? As a matter of pure arithmetic, missing jobs imply one of two things, a huge rise in unemployment, or a huge fall in the labor force participation rate. Can you think of a third factor?”
How about faulty predictions by the BLS? How did the BLS come up with its forecast of 22 million new jobs that Tankersley puts so much faith in? They simply projected from past historical trends. (As they say, past performance is no….)
Tankersley wrote:
“Blinded by low unemployment, lawmakers and economists overlooked two crucial warning signs of the nation’s deteriorating economic health. One was the percentage of working-aged men””the traditional backbone of the U.S. labor force””who held a job. The other was the number of jobs being created each month. Throughout the 2000s, both numbers nose-dived.”
One of these things is true, the other is not, can you guess which is which?
Job creation did slow dramatically during the 2000s but the percentage of working aged men who held a job merely followed a secular trend that dates back to the ealiest BLS records. Adult male participation in the work force has fallen steady as a drum at roughly a 0.23% per year rate since 1948.
What did change? The great decades long boom in female participation in the labor force finally leveled off at just over 60% in 1997. My guess is that it just reached its saturation point.
So there’s not much of a mystery here unless you insist on finding one. Cowen and Kling apparently think that this shows that there were structural problems in the labor force even before 2008. (Which reminds me of the Abraham Maslow quote, “It is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.”)
I, on the other hand, looked at the data from the 2000-2008 and came away with the the exact opposite conclusion. The problem was a modestly slack labor market not a tight one (the unemployment rate doesn’t tell the whole story). You can’t have structural problems with the labor market and still have falling unit labor costs and rising corporate profits. It just doesn’t compute (labor costs should rise).
So we’re back where we started: insufficient AD. (IMO, Cowen and Kling have a case of “structural labor force problems” on the brain.)
21. January 2011 at 13:18
Population growth looks like it was about 1% annually from 1970-1990, then about 1.1% in 1990-2000, then 0.9% in 2000-2010 (those numbers work). If we assumed population growth continued at 1.1% for the past decade and that LFP went up on trend to about 70% instead of down to 66.5%, I calculate the difference at… 15 million.
21. January 2011 at 15:14
Third factor: Slower new firm creation. SarBox and FinReg really hurt new form creation and startup, which disproportionately affects job creation.
21. January 2011 at 15:28
bah, firm not form.
21. January 2011 at 15:32
Doc Merlin,
Actually, Sarbox and FinReg probably increased “form creation”. 🙂
21. January 2011 at 17:03
Philo, Yes, I mentioned how Mexicans who were undocumented were returning to Mexico. But that’s probably not enough.
Contemplationist. It says the inflow has fallen from 750,000 a year to 550,000 a year. But elsewhere I read the drop in immigration was bigger. That could be a factor.
Mikko, But that’s a weird assumption to make, especially given that LFPR didn’t increase between 1990 and 1996. Why would they make that assumption without any good reason?
Mark, I don’t agree with your view of Kling and Cowen. Cowen in particular is one of the most open-minded people in the world. But otherwise your comment is excellent, much better than my post. You should be a blogger.
Dirk, Good observation. If that’s all it was, it makes me more confident that my criticism is accurate.
Doc Merlin, No, that’s not a third factor, it would affect one of the two I mentioned. Also, I doubt it would have a dramatic effect, but that’s just a hunch on my part.
21. January 2011 at 17:23
I’d like to riff a bit off Doc’s thing on firm creation. In particular the costs we’ve seen added via regulatory capture and what I consider to be the damage to Local Wealth, who have taken it on the chin TWICE with both FinReg and now with TARP bailout.
The JOLTS data is still new (since 2000) and I’m going to make a leap that association = causation:
http://www.bls.gov/opub/mlr/2010/05/art3full.pdf
When you look at numbers on job openings and separations – and then specifically at types of separations (quits vs. fires):
We gain the most jobs when we see the greatest turnover.
We gain the most jobs when we have the most quits.
Now I realize that this isn’t prove able, but I think this an example where macro guys screw up their aggregates.
I think that IF you want to create jobs, you just have to create the kind on economy where people QUIT.
And to me that’s an easy economy to create.
First, Local Wealth creates all the new jobs. Even in the tech space, where just 10 years ago we have a institutional VC world, we’re seeing more and more Local Wealth (Angels and super Angels) who basically operate as guys investing in themselves (investing in local companies they can sit on top of) and just paying themselves a salary and declaring profits as capital gains.
BUT, in the real SMB space – the top 2% of SMB make +50% of the SMB revenue… and these guys are the 7M millionaries next door. The stats on them are:
1. they under consume for most of their life.
2. they make multiple local SMB investments – often times with different local partners.
And from this soup of SMBs – all the fast growing new job businesses come from here. Like college football players who might go pro, we should treat them all as the BigMen on campus.
I think we ramp up QUITS faster than anytime since the late 1990’s if we make a single changes to the tax code:
Let all SMB owners pay themselves a single salary and move profits from one SMB to another without ever declaring income – give them a rule like the IRS 1031 on real estate, if you make gains, and put into another building in 3-9 months, you pay no taxes.
And sure there should be no corporate taxes, but I kinda LIKE this changes just for SMBs:
1. It is faster politically.
2. It makes up for all the regulatory capture that the Fortune 1000 foist on us from DC.
I think this would put hungry piranha in the water, and they’d quickly chew through flabby lazy asses of the Fortune 1000 in one year’s time.
We need to stop saying, “when times are good, there are quits”
And we need to ask, “what can cause quits tomorrow?”
This question clarifies to me the problem with jobs in this country, we need the guys most likely to take risks to take risks, and letting them keep all their profits if they take more risks is the obvious solution.
21. January 2011 at 21:24
Also note that the stock market was much higher in 2000 than it is today. You may not believe in bubbles, but it isn’t hard for most people to accept that predictions made for the future of the economy in 2000 were irrationally exuberant. 2000 was the year of magical thinking.
22. January 2011 at 10:24
Morgan, yes, but that doesn’t show there are missing jobs, it just explains what might have caused them.
Dirk, Good point.
22. January 2011 at 11:38
As a matter of pure arithmetic, missing jobs imply one of two things, a huge rise in unemployment, or a huge fall in the labor force participation rate. Can you think of a third factor?
I thought that was an interesting question, so I went back to the BLS projections to see what a third factor might be. Fortunately, the BLS employment projections are very well documented — they run a series of articles every other year in the November issue of the Monthly Labor Review. By comparing the articles describing the 1998-2008 projection that was published in November 1999 with the articles published in November 2009 (giving the actual history for 1998″”2008 and the projection for 2018), I was able to solve the puzzle. The “third factor” is multiple job-holding, ie, second jobs.
We need to keep in mind that BLS gets its employment numbers from two sources–a household survey and a payroll survey. The household survey tells us the number of people in the labor force, number of people employed, and the unemployment rate, while the payroll survey tells us the number of wage and salary employees appearing on employer payrolls. The main conceptual differences are that (a) the payroll survey is really counting jobs rather than people (that is, if a person hold two jobs with different employers, they will be counted twice), and (b) the household survey covers the self-employed, who are missed by the payroll survey.
Anyway, I compared the 1998-2008 BLS projections with what actually took place as reported in the November 2009 articles. Their projection of labor force growth was actually accurate (they projected an increase of 16.9 million, compared to an actual increase of 16.6 million). This was the result of offsetting errors–as Scott noted, they missed the slight decline in labor force participation rate (they projected an increase of 0.5 percentage point, whereas the participation actually declined from 67.1% to 66.0%), but this was offset by an underestimate of the growth in working age population, which they assumed would grow 11.2% compared to actual growth of 13.9%).
For persons employed (“civilian household employment”), the projection assumed a 4.7% unemployment rate for 2008; the actual unemployment rate was 5.8%, so their estimate of the growth in persons employed was about 2 million low (13.9 million actual versus 15.8 million projected).
Their largest error, however, was that they projected the number of jobs to grow faster than the number of people employed. Their projection for “total employment” (which is defined as essentially the number of jobs from the payroll survey plus the number self employed from the household survey) was growth of 14.4% or 20.3 million jobs over the decade, compared to 12.0% or 15.8 million for persons employed. In fact, total employment increased more slowly (7.3% or 10.3 million jobs) than persons employed (10.6% or 13.9 million persons). In other words, rather than adding about 4 1/2 million second jobs from 1998 to 2008, the economy shed about 2 1/2 million. (Caveat–there may be reasons for the differences between the household survey and the payroll survey other than second jobs, so these figures need to be treated with caution.)
Frankly, BLS’s projection that second jobs would grow at a faster rate than persons employed seems a little strange, but that had been the trend during the 1990s, so I guess they were just extrapolating the recent trend. But prior to the 1990s, the number of jobs and number of persons employed had generally trended together, so it shouldn’t be too surprising that the 1990s divergence didn’t persist.
We don’t know a lot about trends in second jobs. I think when people talk about shortage of jobs, I think they are generally thinking of first jobs, so I think the numbers cited by the National Journal article, which were largely driven by a change in the trend of second jobs, are a bit misleading to the whole discussion.
22. January 2011 at 15:14
Brent,
Excellent research. I enjoyed your comment. Very illuminating.
22. January 2011 at 21:30
I noticed that I made a small mistake above – in the third from last paragraph where it says “from 1998 to 2008, the economy shed about 2 1/2 million,” it should have been “3 1/2 million.”
23. January 2011 at 09:04
[…] Nick Rowe talks about the fact that housing transaction volume is higher when prices are rising; Scott Sumner’s latest attempted swindle; . . […]
23. January 2011 at 19:42
Brent, Thanks for all that research. I added a quotation from your smaller comment later to one of my recent posts.
3. March 2011 at 16:08
[…] TheMoneyIllusion » A missing explanation for why the National …Jan 21, 2011 … A missing explanation for why the National Journal says we have missing … to a National Journal article that implied America had lost (or … […]