We mentioned Labor Force Participation Rate over the weekend. It is worth exploring a bit further.
One of the lesser noted aspects of the NFP report Friday was the drop in the Labor Participation Rates. It is the percentage of the total population that is either working or potentially working. In the US, it is about ~170 million people, and excludes children, retirees, non-employed (work at home moms, etc.). That Labor Pool measure peaked in 2000 at just under 67.4%. Following the market crash and recession, it subsequently fell about 1.5% to about 65.8%. This accounted in no small part for the falling Unemployment rate from 2001 – 05.
Now, a 1.5% or so drop doesn’t sound like a lot, but remember there
are 143 million workers in the US. That drop equals about 3 million
people. These are folks who are willing to take a full time job, have
been unable to find work, and have exhausted their unemploment
benefits.
The Labor pool drop appeared to have bottomed and reversed it self late 2005 (see chart above). The 5 year downtrend channel was broken, and a new uptrend — higher highs and higher lows — was beginning.
Until recently.
It has since started heading lower again. They do not count in the "official" Unemployment Rate statistics. However, BLS actually does measure these folks in their "augmented unemployment rate" — the jobless people who aren’t counted among the officially unemployed. That measure is 7.4%.
There is a good definition of this here. The longer term chart below parallels the unemployment rate — but from higher levels.
Some people have asked if this 7.4% level is a good or a bad thing. That qualitiative description is less relevant to me than the quantitative cobnclusion that the 4.5% Unemployment Rate is very misleading.
David Altig at Macroblog thinks it is mostly due to secular changes, and I do not disagree. But for policy makers — such as the Federal Reserve — the belief that the labor market is very tight with no slack is clearly belied by the data.
The significance: It may have an impact as to whether they have the room to cut rates or not.
I guess the tile “Civilian Labor Force” means private business employment and discludes government employees.
I wonder in Econ301 how the government worker / taxpayer supported worker gets weighted with mining / manufacturing / service industries?
My point is the balance of it all.
Mining digs up 100Million year old inventory, so thats new currency.
Manufacturing blends mined products, so thats new currency.
Services blends the above and becomes the middle man, and our consumerism supports this need of ours.
That leaves government employees and the money industry call it Capitals and WallStreet. These services extract without creating. Both masses are needed in our current system.
My point is the balance in truth presentation / representation.
A revolving dollar is a working dollar in any of the major classifications. My problem is globalism, because I’m not moving anywhere, except to dust.
On the augmented chart the 7.4% rate appears to be approaching the low of somewhere around 7% in mid 2000. In 1998 on the augmented chart the rate appears to have gotten to around 8%.
Wouldn’t that suggest that the labor force is similiarly tight to the 1999 timeframe regardless of which measure you use? Or is there something else that plays into it that is not clear on these charts?
This accounted in no small part for the falling Unemployment rate from 2001 – 05.
I think unemployemt rose until 2003. Hence all the gripes about the “jobless recovery” and contentions that we were really in a recession until 2003.
~~~
BR: You are correct — Unemployment rate bottomed in 2001, spiked up until 2003, and then moved downwards over the past 4 years . . .
See this chart:
http://www.economagic.com/chartg/feddal/ru.gif
WASHINGTON (MarketWatch) – Businesses boosted their capital spending in April at a faster pace in April than previously reported, the Commerce Department said Monday.
Orders for core capital equipment goods rose 2.1% in April, revised up from 1.2% reported two weeks ago, the government said. Shipments of core capital goods – which exclude both defense goods and civilian aircraft – increased 1%, ahead of the 0.7% previously reported.
Combined with strong March data, the report shows a faster recovery in business spending and factory output, confirming other indicators of strength, such as the Institute for Supply Management index and industrial output data from the Federal Reserve.
You said:
“Now, a 1.5% or so drop doesn’t sound like a lot, but remember there are 143 million workers in the US. That drop equals about 3 million people. These are folks who are willing to take a full time job, have been unable to find work, and have exhausted their unemploment benefits.”
This is factually wrong. Whether or not someone is collecting unemployment benefits or not has no bearing on whether they are counted as unemployed. You are counted as unemployed if you have spent time looking for a job over the past month and not found one. Anyone who is willing to take a job and is looking, is counted as part of the labor force.
Secondly, the decline in labor force participation in the economy at large is largely due to a decline participation among teenagers. There are a lot of possible reasons for this but it is hardly a reason for worry about the broad economy.
daisycolorado: You are right that the BLS unemployment definition is independent from benefits, but there is a credible indirect link — in order to continue receiving full benefits or any at all, you have to declare that you are making job-search efforts, and presumably may be requested to supply supporting evidence.
This creates an incentive (of a strength depending on how much auditing actually happens) to at least keep up appearances even though you are not confident of finding something. Then there is also the time factor, after 6 months looking when your benefits run out, 6 months have also passed during which you didn’t land a job, or perhaps not even an interview (or fewer, if you didn’t manage to hold your job long enough to restore 6 months eligibility). I don’t know how long your spirits would last.
BR – “But for policy makers — such as the Federal Reserve — the belief that the labor market is very tight with no slack is clearly belied by the data.”
I don’t see where the belief is so belied. Slack implies that if demand increases, resources are available at the margin to satisfy the increase without bidding up prices.
The fed needs to look at the nature of the incremental demand, and the availability resources needed to accomodate that demand. Suppose, for example, our NILF slack is largely underemployed carpenters, but incremental demand is coming in the form of ICU nurses. In the long term, carpenters may become ICU nurses, but in the timeframes the fed thinks about, the incremental demand for nurses can’t be met by the resources available, and the likely result is a bidding up of the price of trained ICU nurses.
I don’t know if this is the case now, but the fed is likely looking beyond the aggregate NILF number to determine whether the “slack” is likely to mitigate near-term wage price pressures.
High-level definitions from the BLS household survey, Technical Note:
From section A-12, Alternative measures of labor underutilization:
The BLS/CES websites have material elaborating more on details of what the meanings of “job”, “specific efforts”, “available for work” etc. are, but it takes some searching and I’m a bit in a rush now.
manhattan guy-
Key phrase that you seem to continually miss in just posting snippets of articles that support your view.
“revised up from 1.2% reported two weeks ago, the government said”
All these revisions upward shows they have absolutley no clue as to what they are in the first place.
How come they are never revised down??
Since they are engaged in underreporting do you think that may be they could be overreporting in the first place???
Of course you don’t because you have no opinion or insight to any of these bull articles that you post.
Where’s Fred with the “Capex will save us” arguement as I totally expected that post from him.
Ciao
MS
Barry,
I’d love to see some work on the “hidden” job market…the off the books work that gets into none of these stats. I this it is HUGE. How many people out there are either living off or augmenting their income by ventures like “selling stuff” on Ebay?
Manhattanguy…ignore MS, and his drivel.
These labor force participation rate discussions are quite frankly just silly. Are there less people in the labor pool than in the boom years? Sure. Does that mean the labor market is slack? No.
First and most importantly unemployment is the stat that we have used for a century. Just like “approval rating of the president” then true meaning is in its relative performance to past values. Not trying to measure something absolute today.
Yet, the fact actually is that unemployment does measure what we want to measure. How many people are looking for a job but have not found it. That is the ratio of job seekers to job holders. There are lots of people who would take a job if it paid their asking wage and was handed to them.
I am sure Manhattan is full of socialites who would take a job as an international supermodel if it were just handed to them.
More relevantly there are probably tens of thousands of stay-at-home moms who would take a job if it offered good pay and totally flexible hours.
However, an increase in the availability of both of those jobs would represent an effective increase in the wage. Most socialites are not very “productive” as models and flextime workers are not as productive as 9-5ers. If the job market were willing to pay them the prevailing wage then it would mean that unit labor costs are rising.
So an increase in the job participation rate goes hand in hand with increases in unit labor costs. That is precisley what the FED is concerned about.
Uh, Karl… any references to back up your claim that flextime workers are not as productive as 9-5ers?
I find otherwise:
In general, these programs have resulted in reduced turnover and absenteeism, higher employee morale and productivity, and improved worker well-being (Gale, 2001; Gill, 1998; Lucas & Heady, 2002).
Gale, S. (2001). Formalized flextime. Workforce, February, 39-42.
Gill, B. (1998). Flextime benefits employees and employers. American Printer, 5, 70.
Lucas, J.L., & Heady, R.B. (2002). Flextime commuters and their driver stress, feeling of time urgency, and commute satisfaction. Journal of Business and Psychology, 16 (4), 565-571.
MS – I post these *bullish* articles because BR ignores to post them (ahh we all know he’s a bear). When are you going to learn that investors pay for future earnings and growth. You base your bearish opinions on what happened the last few years. And enough with all the negative posts around real estate. You sound like a broken record. That was news from two years ago.
While I agree that there is a credit bubble going on, in my opinion S&P is still undervalued. We still see some upside left in the market. I see Dow to hit 14000 and S&P 500 to hit 1550 before year end.
Combined with strong March data, the report shows a faster recovery in business spending and factory output, confirming other indicators of strength, such as the Institute for Supply Management index and industrial output data from the Federal Reserve.
ManhattanGuy,
I think the rise in the stock market is stimulating business spending. That said, the consumer spending numbers for the past two months show weakening (real PCE 0.0% and 0.2% in March and April). In the short-term this keeps the bear case alive IMO – consumer spending now accounts for almost 72% of the GDP. If the weakeness in consumer spending accelerates that will create a “classic” housing-led recession where a drop in consumer spending kicks the legs out from under business spending. These numbers need to pick up (or be revised up) in coming months or a 2007 (or 2008) recession is a a real possibility.
Karl Smith: There are substantial caveats to your claim about relative comparisons in unemployment (over the long time span you mention), which are (1) the definition and measurement/estimation of unemployment has not remained constant, with a bias towards defining/leaving out more of those who are arguably “not working” involuntarily, and (2) the labor market has changed structurally, converting a good number of previously “9-5” fulltime positions to part-time, contract, or temp jobs, for a bias towards more jobs for the same amount of work.
And the meaning of a job has changed too — you will no longer get the same standard of living/quality of life per household income as a number of decades ago.
Wouldn’t that suggest that the labor force is similiarly tight to the 1999 timeframe regardless of which measure you use? Or is there something else that plays into it that is not clear on these charts?
Who cares. The BLS has been corrupted and changed methods in 2001. Hence, augmented, not augmented, it doesn’t matter, the same “mistakes” in procedure are still being made yet we are told they are still the same?. The economy is running at 1995 speed, not 1999.
FWIW, Kennedy-Clinton had the UE rate at 5.4% for May.
I have learned today the following:
The BLS’s BED unit puts out their quarterly report always during the middle month of each quarter.
Those months are always Feb, May, Aug and Nov. They do not announce the date in advance (you can sign up for an email notice when BED is released) and there is no specific key day of the week or month that they use, as in the BLS NFP which is “first Friday” typically.
I’ve looked back and releases have been made randomly on different days of the month, for example on the 3rd of a month and on the 18th, so I don’t suppose we can even say they would be mid-month for sure.
They are always released exactly 8 months and possibly a few days from the close of a quarter being reported. For example, we just had Q3 2006 reported and it’s 8 months from Sep to May.
Obviously the next release will be in Aug.
Uh, Karl… any references to back up your claim that flextime workers are not as productive as 9-5ers?
The existence of 9-5 jobs. Here I don’t mean necessarily 9am – 5pm but stated start and stop times that are enforced by management. Unless your arguement is that people en masse prefer stated schedules its hard to imagine why people would work for less money in a situation that was less preferable. Now it may be true that:
1) There are increasing opporuntities for flextime work that is more productive and in those industries we can expect the traditional job to disappear. Here in RTP most IBM workers I know do not have to go into work and so the model is being replaced.
2) The workers who are getting flextime are inherently more productive and are choosing to spend some of their productivity on quality of life.
I am in no way saying that flextime is a bad idea, I am simply saying it stretches reason to expect that Wal-Mart can opperate as effeciently by telling it works to “come in when its convient for you”
Importantly, unemployment at the high skill end is practically non-extistant, so its the Wal-Mart worker that we are thinking about.
the definition and measurement/estimation of unemployment has not remained constant, with a bias towards defining/leaving out more of those who are arguably “not working” involuntarily
As far as I know the headline number still comes from the following two questions.
1) Are you employed.
2) If no, have you looked for a job.
Now I am by no means arguing that there haven’t been structural changes in employment. However, measuring that is not what this number does, nor is it what this number is designed to do.
You simply want a different statistic. Indeed, you probably want an index that no one has developed yet. However, we should not be popping another number into back of the envelope Phillips Curve-esque analysis simply because headline unemployment isn’t everything to everyone.
Karl Smith: As you are quoting Wal-Mart as an example for fixed work schedules, this is not primarily about “productivity” but the fact that the job requires predictable physical presence over specific time periods, which is also the reason that bathroom breaks are regulated. Of course, in many work places productivity depends on everybody being in the same place at the same time as there are strong interdependencies either of communication or interaction between work units.