The Real Unemployment Statistics, or what the government doesn’t want you to know about the economy

What I’m going to say here will seem to go against everything you’ve heard in the media and from the government for the past several years on the state of the economy.  If you actually read through this you’re either going to dismiss it as foolish and misinformed, or, I hope, search out my references and confirm for yourself what I’ve been tracking for years, but haven’t bothered to write about until now.

Fair warning, this is probably going to be one of my driest posts yet.   If you read on you will be subjected to graphs, charts and data and other material that will seem like it came from that ECON201 class you hated so much in college.  I’ll try to keep it to graphs, and pictures rather than tables of data so this is a little more entertaining, but bear with me.  This is tedious stuff that only an economist or someone suffering from sleep deprivation and insomnia would look at.  And I think that’s why, when you read my findings, the government thinks they can get away with this.

For those of you who don't want to read the entire post, here's the punchline

Executive Summary

Since May 2009 the BLS has consistently and significantly understated the true U-3 unemployment by at least 3% versus what it should be, this gap has increased by about 1% since about September 2010 to almost 4% and appears to be increasing.

Employment Improves Dramatically! 

Economy on its way to recovery

Much has been made of the improvement in state of employment over the past several months.  Headlines reporting the dip in unemployment have driven up the markets, Democrats have patted each other on the back for turning the tide, Republicans have gnashed their teeth, wondering if the economy will continue to turn for the better prior to fall elections.  But has it really?  This is a tremendously important issue because, these data (the Unemployment 3 or U3 report issued by Bureau of Labor Statistics) is the benchmark whereby the government policy makers, Congress, markets and industry gauge the health and prospects of the economy. 

Friday (February 17^th) the financial markets (Dow Jones Industrial Average (DJIA)) closed close to a bellwether mark – 13,000, up 2,400 points in the last 4 months, returning to levels last seen before the 2008 crash.  (The LA Times: Dow nears psychological milestone: 13,000)

A major driver of this enormous gain has been euphoria over the unemployment situation.   But what if unemployment gains have really been more modest that reported?  What would the markets have done?  And more importantly , what will happen once they discover their mistake? 

On a personal note, can your 401(k) take another hit like it did in 2008?  Hmmm.  Let’s take a closer look

Labor Statistics, or (yawn) I forgot my homework at home professor

Using “seasonally adjusted data,” the Bureau of Labor Statistics (BLS) tracks and reports on labor statistics.  It’s the primary source of labor information for government policy makers, economists, investors, and political wonks who want to twist information to say whatever their preconceived notions (be they left or right) tell them they should say.

On February 3, 2012 the BLS reported that:

 “Nonfarm payroll employment rose by 243,000 in January, and the unemployment rate decreased to 8.3 percent. Job growth was widespread, with large gains in professional and business services, leisure and hospitality, and manufacturing.  Government employment changed little over the month.”

 http://www.bls.gov/news.release/pdf/empsit.pdf

An impressive result, bringing unemployment down to its lowest level since it peaked in October 2009 at 10.0%, and yet still 3% higher than when President Obama took office in January 2008 (5.0%).  Or was it?

To understand this we need to do a little math and follow how the BLS calculates unemployment.  The data I’ll be using for this discussion are the “seasonally adjusted data” which are the data the BLS uses for its reports.  Basically they look at labor force seasonality (say lots of temporary employment around Christmas as retailers hire temporary workers to work the shops and then cut them after, in the summer, when agriculture employs migrant, oops, I meant to say temporary laborers to harvest the field and so forth) and adjust (read “tweak”) the numbers to avoid noisy fluctuations in the numbers. 

While the impact is of making this seasonal adjustment is significant, these are the numbers they use, and the numbers the market reacts to.  For example in January 2008 the seasonally adjusted population was 154,075,000 and the unadjusted population 152,858,000, or a  .8% increase in the denominator of the unemployment calculation (but I get ahead of myself).

So, on to understanding what the numbers tell us.

Total Population

The BLS defines the Total Population as “the sum of employed and unemployed persons. The labor force participation rate is the labor force as a percent of the civilian noninstitutional population.”  Hmmm.  What does that mean?  Well, they remove from consideration people living outside the country,  disabled, or either too young or too old to work.

In February 2012, the BLS reported the Total Population as 242,269,000, which means that about 70 million Americans are not part of the BLS statistics.  This is the first interesting place the data start to unravel.  As the economy worsens, people’s retirement programs disappear, costs increase beyond Social Security payments, increasingly the elderly are returning to work.  So you have to question any methodology that unilaterally excludes the aged from the base population.

Working Population

This is, well, what it says.  The BLS defines Percent of Population working as those individuals who are “active in the labor market”.  These are people who are either working full-time jobs, part-time jobs, or out of work and actively seeking employment with certain exclusions including workers who have just given up looking for a job (disaffected labor).  (We’ll speak more on this later on)  The BLS defines Working Population as the civilian, noninstitutional population, excluding military, and those incarcerated or otherwise institutionalized.    (See my footnote at the end of this entry about incarceration in the United States, and its impact on Labor Statistics.)  For February 2012 the BLS reported the Seasonally Adjusted Working Population of 154,395,000.

Working Population Percentage

This is used to report what is essentially the percent of the civilian workforce that is working or actively seeking employment.  It’s calculated as

Working Population (Seasonally Adjusted) / Total Population

For January 2012 the Seasonally Adjusted Working Population Percentage was 63.73% (154,395,000 / 242,269,000).

Whoa, this is a little much

Let’s stop here for a second and take a breather.  Let’s look at what these two numbers look like over the last ten years based on BLS data (for all the graphs and figures see the BLS website where you can download the source data and play with it to your hearts content.)

In January 2002 the BLS reported the Total Population as 216,506,000 and the seasonally adjusted Working Population as 143,883,000.  Let’s us this as the starting point (set both to 0) and look at how they grew or shrank over the last decade:

Here we can see that the Total Working Population grew by about 25 Million in the last decade, while the Seasonally Adjusted Workforce grew by about 10 Million, basically leveling off in the summer of 2007, about a year before the market crash of 2008.  Hmmm.  What’s up with that?  Have we had a massive egress from the US?  Are there 15 Million more people in the military or in prison?  No.  This means that a large number of normally employable people have permanently left the workforce. 

Let’s look a little closer.

Let’s say (for lack of a better starting point) that January 2002 represents the “normal” full participation in the workforce.  What would that be? 

Working Population (Seasonally Adjusted) / Total Population =

                        143,883,000 / 216,506,000 = 66.46%

So basically, all other things being equal, there hasn’t been a tremendous change in the total number of people that COULD be working.  Sound like a reasonable assumption?  OK, then let’s look at what the above graph would look like of we kept the workforce at that rate.

If you follow the purple line in the chart above, then you can see that about 6.6 million people have left the active workforce in the last 10 years.  Wow.  Let’s put this in perspective .

How is unemployment calculated?

This calculation is relatively simple.  BLS defines the U-3 rate (people with full time jobs) as one minus the seasonally adjusted number of people employed divided by the seasonally adjusted working population.   (Note – I won’t address the other factors of a weak labor market this the government reports as U-4 through U-6 as this topic has been addressed extensively by places such as the Economic Policy Institute.

So for January 2012 a the BLS reported:

(1 – (141,637,000 / 154,395,000)) = 1 - .9174 = 8.26%, which they report as 8.3%.

What’s the impact of all those people who “left the workforce?”

Well, these are people, who would normally be working, yet somehow have been excluded from the numbers.  To consider them, you’d have to add them back into the denominator (the working population).  What would that do?  In January 2012 the impact of this factor is about 6,609,000, which would increase the working population to 161,004,000. 

So what’s the unemployment rate if we incorporate these 6 Million people back into the equation?

              (1 – (141,637,000 / 161,004,000) = 1 - .8797 = 12.03%.

Holy Crap!  Unemployment (U-3) is really about 4% higher than the government is reporting it!

Let’s look at how this trends over time versus reported unemployment once again setting January 2002 to zero:

Key Takeaway 

Here you can see that since May 2009 the BLS has consistently and significantly understated the true U-3 unemployment (the red line) by at least 3% versus what it should be (the blue line), and that the gap is between these two calculations has  increased by about 1% since about September 2010 and appears to be increasing.

Wow.

What I have to ask is why is no one reporting on this?  Neither the liberals or the conservatives seem to have picked up on it. 

I’m not surprised you don’t hear about this in the mainstream media – hey, it’s hard work to explain this, even harder to understand.  It took me about 2 hours to analyze the BLS data and create the graphs you’re looking at and another 4 hours to write this.  But you’d think that policy makers or Wall Street traders or bankers would be seeing this and screaming bloody blue murder.

But they’re not.

Hmmm.  Perhaps that’s a topic for another day, but I grow weary and you’ve probably lost interest.

If you made it this far, congratulations!

Until next time – If you’re not pissed off, then you haven’t been paying attention.

Incarceration in the United States, an aside

 In 2009 (the latest date for which data are available from the Bureau of  Justice Statistics), there were 2,284,900 prisoners in federal prisons and state jails. 

Source: http://bjs.ojp.usdoj.gov/content/pub/pdf/cpus09.pdf

The US Census estimate for 2009 was 305,529,237 (US News and World Report), which puts the US incarceration rate at about 0.75%  For 2011, provided the rate remained constant, this would have meant about 2.3 million Americans would have been in prison or jail.  To put this in context, the US has about 4 times the average global rate of incarceration, putting us at the highest rate in the world.  (See The National Council on Crime and Delinquency)

Once again We’re Number One! Go USA.