Official Unemployment Numbers Understate the Problem

Mike Shedlock (Mish) has produced an excellent graph in his article, "The Decade of Lost Jobs", which can be read in full here. (Hat tip to Phil's Stock World.com/.) Mish's graph is reproduced below and is followed by my discussion of some of the factors behind the trend shown. The conclusions may surprise you.

In looking into what's behind Mish's work, I chose to look at the total employment numbers from the DOL. One of Mish's conclusion is that more than 2/3 of the meager job growth of the past 10 years has been in government jobs, not in the private sector. I have nothing to add to that conclusion.

The following table was constructed (by me) with data from the U.S. Dept of Labor (DOL), using Seasonally Adjusted data:

In the past ten years, the following has occurred (from data in table above):

• The civilian population over 16 has grown by 28 million
• The labor force has grown by 16 million
• The number of those not in the labor force has grown by 12 million
• The number of jobs has grown by 7.3 million

Although the number of employed has increased over the 10 years by 7,259, the number of unemployed has increased by 8,715. This constitutes a relative job shrinkage of 1.5 million by subtracting the increase in the number unemployed from the increase in the number employed (7,259 - 8,715 = -1,456). If you add these two numbers together (instead of subtracting), the total is 15,794, the exact increase in the labor force reported by the DOL, so the data is internally self-consistent.

The growth in the number not in the labor force results from an increase in the student age population still in school, increase in the number of retirees and increase in the number of discouraged unemployed no longer looking for work. So, the past ten years has seen an increase in working age people not in the labor force (12 million), almost as large as the increase in the number of people in the labor force (16 million) and a decrease in the relative number of jobs of 1.5 million.

This is not a pretty picture: There was a gain of 11.5% in the labor force, compared to a gain of 17.6% in the total not in the labor force. During this period, the population grew from 272.4 million to 306.5 million (U.S. Census Bureau), or 12.5%. The ten-year population growth rate was greater than the growth rate in the labor force and much less than the growth rate for those not in the labor force. These numbers are summarized in the following table (Seasonally Adjusted data):

Currently the unemployment rate is at 9.4%. Many think it will go higher before it goes lower, but what would things look like today if the unemployment rate were to return to 5%? Those numbers are summarized in the following table, using Seasonally Adjusted Data:

Unemployment of 5% is generally considered to be the equilibrium point which is considered virtually full employment. If we were at that point today, the 7.3 million jobs gained during the past ten years would be increased by 6.8 million to 14.1 million jobs gained. That number should be compared to the increase in the labor force since 1999 of 15.5 million. This (14.1 million jobs gained) is close to the expected number 15.2 million (95% of 15.974), but is still 1.1 million short. This close to the 1.5 million found earlier for the difference between the increase in the number unemployed and the increase in the number employed over the past 10 years. It seems to me that these two numbers should be the same and I have not rationalized why they are different.

At 9.4% unemployment, we are almost 8 million jobs short of keeping up with growth in the labor force. At "full" employment (unemployment of 5%), we would still be over a million jobs short. These estimates are influenced by two facts: (1) more people have lost jobs (or failed to find employment upon entering the labor force) in the 10 year period than have gained employment and (2) the number of people not in the labor force is increasing at a much faster rate than the number in the labor force.

Another factor that is not included in this discussion is the effect of slack in employment due to a large number of people working part-time for economic reasons (Ie, not by choice). There has been a large rise in involuntary part-time employment over the past 10 years. In a previous article (here) I have estimated that the implied number of unemployed if you count 40 hours per week as one person employed is possibly as high as 39 million. While this estimate is certainly too high (I am working on a better calculation), the final estimate will be probably be in the range 20 to 30 million, much larger than the official unemployment total of 14 million.

The large increase in involuntary part-time employment is likely hiding a more severe distress in the labor market than is portrayed by U-3. Even U-6 may understate this distortion. More will come on analysis in this area in a later article after more research is completed.

The rest of the world is also seeing a dramatic increase in part-time employment, as reported by the Financial Times here.

In conclusion, the current unemployment rate (whether U-3 or U-6) is not properly accounting for some serious problems. These are:

1. Over the last 10 years, the number of people not in the labor force has grown proportionally faster than the number in the labor force.
2. Over the past 10 years, the number of unemployed has grown more than the number of employed.
3. The impact of growth in involuntary part-time employment is not properly recognized in the official DOL unemployment rate.

Comments

[markfl:]

Your post helps illustrate the recovery problems we face in 2010. In Jan. 2002 total employment was 133.4 million. In Dec. it stood at 133.9 million. The economy created less than 50,000 jobs per mo. The unemp. rate increased from 5.6 to 6% (seasonally adjusted U-3), with the rate among blacks at a whopping 11.5. U-6 (the national underemployment rate) hit a non-adjusted 9.6%. Source: BLS.

GDP grew at 1.6 in 2002 and created basically 500,000 jobs--as many as we've lost in one month.

Jul 02 05:15 AM

[Steven Hansen:]

John, since you are reviewing the same month of the year, part of the distortion could be caused by changing seasonal adjustment factors. i am having internet problems today and cannot get into bls to validate what i am thinking by looking at the nsa data. another potential reason is that there has been an adjustment for the 2000 census. either of these could account for the subtle data distortions you discussed.

the growing issue of declining full employment is being swept under the carpet by each presidential administration. there has been no concerted effort by any president i can remember to do more than just talk about jobs.

our free trade agreements are effectively lowering the costs of goods and services while exporting jobs. no solution is being pursued as economists believe efficient cost structure is more important then jobs. no president, including our current one, has had the balls to stand up and say "jobs are more important than trade".

Jul 02 05:25 AM

[User 353732:]

1. As the Parasitic Econmy expands inexorably , so the Productive Economy shrinks
2. The Productive Economy must shrink faster than the growth in the Parasitic Economy because the non -productive overhead, waste, fraud of the latter is much higher and, of course, value added is much lower than the former
3. The more national economic, intellectual and cultural resources are allocated to the Parasitic Economy, the more there is a shift of resources from higher value to lower value activities and transactions and hence a degradation in the aggregate wealth producing capacity and global competitive profile of our econonmy
4. Consequently, the ability of the economy to generate NET, new, jobs deteriorates appreciably: this is manifested by more active unemployment(no job; but looking); passive unemployment(no job; given up looking); underemployment(want to work more but cannot get the work); degraded employment(same job but less real income for a variety of reasons) ; inferior time allocation that suppresses the unemployment level( people who are "students" because they cannot find jobs, for a variety of reasons including lack of relevant skills, attitudes and knowledge) and crime( go to jail; disappear from the ranks of the labor force; live off others).
5. Both new entrants to the labor force and those on the margins of society , by choice or circumstance suffer the most as the Parasitic Economy expands: for these people , the lack of opportunity and the pernicious misallocation of national resources is a very high and very cruel tax. They pay for handouts and Government"compassion" with despair, humilation and alienation from both democracy and free enterprise)indeed from liberty itself).

Jul 02 06:08 AM

[Dave Wrixon:]

Don't worry, a solution is at hand. If you only count those in Government Jobs as being in the labour force then the figures are much improved. LOL

Jul 02 07:13 AM

[CautiousInv...:]

As always, interesting stuff John. I tried to vist the BLS site but it was down in advance of today's release.

When get a chance, though, I want to learn more about the strucure of employment. I am certain that losses in manufacturing have been offset by gains in government, quasi-government and service sectors employment.

Another dimension is the length of the averaage work week and today's report also showed the average work week fell to 33 hours, the lowest level since records began in 1964, from 33.1 hours in May

Jul 02 08:50 AM

[Edward Harr...:]

John, great post! Obviously, anyone willing to take a holistic view and try to make the numbers apples to apples will come to the same conclusion you have done.

I would also add that, in most recoveries, discouraged workers and others not counted in the labor pool start re-entering the labor pool, attracted by the potential for work. What this effectively means is that unemployment surges even after recovery arrives because the labor force participation rate normalizes again.

This suggests an unemployment rate north of 10% is very likely.

Jul 02 08:52 AM

[ain't no fo...:]

Latest report just in and the numbers were not good.

To follow up on John's statements about the surge in part time employment, the average hours worked fell to a flat 33.0 in June, a new record low and continuing an established trend.

It occurs to me that when/if a "recovery" begins, and especially if it is tepid at best, many of the still employed part timers will receive additional hours before companies take on the expensive task of hiring more workers. This would tend to keep many of the unemployed just that - unemployed, putting a drain on the overall economy and especially on cash-strapped states which will have to keep paying unemployment benefits even longer.

There were also another 185,000 net "NEW" jobs magically discovered through the birth/death model in June, bring the past 5 months of net "NEW" jobs to 879,000 - we're in the worst economic disaster since the Great Depression and yet they still concoct these magical new jobs numbers and feed them to the public month after month to keep the REAL statistics lower than they actually are - who in their right mind can really believe we are creating so many net new businesses in this broken economy which have created that many jobs?

I mean c'mon guys, if you're gonna LIE, you need to at least make it believable!

Jul 02 09:17 AM

[Pat C:]

Great objective article. As a novice economist (unemployed and absurd amount of time on my hands), I would enjoy doing some research on the economics of following full-employment (not sure what that is: 5% unemployed, 40 hr. workweek etc) as the primary driver for economic success rather than GDP, capitalism, efficient markets, return on equity etc. etc. which creates so much volatility and waste. It would seem to me that utilizing the full potential of our human resources should be the end game and should be what you "solve for" and not just an "input", result or lagging indicator.

If you or anyone can give me some resources I would sure appreciate it.

Jul 02 09:19 AM

[Lilguy:]

Excellent article! Gives a great boost to the argument that we have been in a secular bear market since the bubble burst in 1999. It also makes it hard to believe that there will be a significant economic recovery anytime soon. In fact, it looks kind of grim.

Jul 02 10:07 AM

[Larry House:]

You are right, John. Bill Gross puts the number of unemployed and partially employed at about 35 mil. That does not lead to a robust economy.

Jul 02 10:30 AM

[John Lounsbury:]

Pat - - -

This is an active research area for me and I expect to publish some more work here on Seeking Alpha and on Real Money at TheStreet.com in the next few days and weeks.


On Jul 02 09:19 AM Pat C wrote:

> Great objective article. As a novice economist (unemployed and absurd
> amount of time on my hands), I would enjoy doing some research on
> the economics of following full-employment (not sure what that is:
> 5% unemployed, 40 hr. workweek etc) as the primary driver for economic
> success rather than GDP, capitalism, efficient markets, return on
> equity etc. etc. which creates so much volatility and waste. It would
> seem to me that utilizing the full potential of our human resources
> should be the end game and should be what you "solve for" and not
> just an "input", result or lagging indicator.
>
> If you or anyone can give me some resources I would sure appreciate
> it.

Jul 02 10:42 AM

[markfl:]

More to the point, average weekly earnings and index of aggregate weekly hours hit a three-month low.

On Jul 02 08:50 AM CautiousInvestor wrote:

> As always, interesting stuff John. I tried to vist the BLS site but
> it was down in advance of today's release.
>
> When get a chance, though, I want to learn more about the strucure
> of employment. I am certain that losses in manufacturing have been
> offset by gains in government, quasi-government and service sectors
> employment.
>
> Another dimension is the length of the averaage work week and today's
> report also showed the average work week fell to 33 hours, the lowest
> level since records began in 1964, from 33.1 hours in May

Jul 02 11:42 AM

[John Galt:]

When the Roman Empire collapsed,

1/3 of the people worked for Caesar ( Government/military etc.)
1/3 of the people didn't work and were supported by the government
1/3 of the people supported the system via Private Industry.

The Empire wasn't always like that, but it evolved to that. Now we see our Baston of Freedom, a country that was founded by people who believed the government should fear the people instead of the people fearing the government turn into that...

America was founded by people who thought government was a "neccesary evil" and that you should "keep fire in the fireplace". To put limits on the government they created a constiution to LIMIT a governments power over the people.

The genius of James Madison INVENTED a bicameral legislature to seperate powers... because the founders FEARED the power of government.

They created a system of checks and balances because they FEARED power, and they didn't want a lot of CHANGE. Change was to be carefully deliberated instead of having 1500 page bills rushed through Congress.

The system is rotting. Rome is burning. It's VERY sad.

Jul 02 01:47 PM

[billddrummer:]

Excellent article and comments.

Jul 03 12:45 AM

[DavyJ:]

Re: "This (14.1 million jobs gained) is close to the expected number 15.2 million (95% of 15.974), but is still 1.1 million short."

95% of 15,974 is 15,175, not 15,200,000, so strike that mystery (and the entire discussion in the paragraph).

And although the statistics are interesting, and represents a lot of work on your part, there isn't anything surprising in your conclusions nor anything not well known. I apologize for being negative, but I did spend an inordinate amount of time digesting your analysis in addition to the above mentioned correction.

Jul 04 10:58 PM

[DavyJ:]

Disregard my comment above. I see a decimal point where before I saw a comma.

Jul 05 12:26 PM

[John Lounsbury:]

DavyJ - - -

Don't feel badly - been there, done that.

To any other readers who might be confused, the data in the tables is thousands of persons and in the text the numbers are expressed as individual persons. Thus 15,175 in a table is 15,175,000 individuals, which rounds to 15.2 million.

This may be confusing, but I've fallen into that trap by reading the DOL data and reports and that is exactly what they do.

Jul 05 02:38 PM

[Leftfield:]

Statistics that are hard for government to fake look far worse. There is no "birth/death adjustment" equivalent that BLS fakes imputed new jobs in industries such as real estate(!), in income tax collections. California is down about 40% I believe. They haven't marked that to magic yet.

Jul 05 04:23 PM