Thursday, September 20, 2012

The "Real" unemployment rate?

Businessweek just ran an article about an increase in Social Security benefits for retirees. The article notes the total number of benefit recipients: 56 million. If we add this to the BLS statistic of 155 million for the labor force, we get 211 million - and there are another 100 million Americans who are not counted. Clearly many of those are children, (75 million under 18 years in 2010, according to the Census Bureau). So now there are an additional 25 million who are not retired, not disabled, and not receiving social security benefits, but also not counted in the labor pool, and thus not counted as unemployed. If we take out 1.6 million or so for our prison population we've cut this down to 23.4 million adults.  Next let's take out 7.8 million full time college students (61% of the 12.8 million college youths) and that brings us down to 16.6 million. 

Another article has come to my attention today.  It discusses stay at home parents and includes census data.  The numbers provided for 2011 are stay-at-home dads make reached "176,000, or 3.4 percent" of stay-at-home parents.  Dividing one by the other gives us a little less than 5.2 million stay at home parents, so let's also drop those out of the uncounted.  16.6 million less 5.2 million means there are 11.4 million  additional adults that are not employed but simply not counted. 

The bottom line is, if these uncounted people were added to the labor force and factored into the unemployment rate, it would be 14.1%, not even counting underemployment.


I've long been adamant that the government statisticians cook their books.  I decided to review the unemployment numbers and see whether I could determine, from their data, how different the numbers should be.  While this is purely an exercise in skepticism, it does provide some interesting insights into the labor market.

First I took a look at the labor participation rate, and noted that it's been in decline since October 2008. This data is used to separate out those who are counted as job-seekers (employed or unemployed), or just not counted.  Now a decline of 3% sounds pretty small, but when you apply it to ~200 million people it turns out to be pretty big - about 6 million. 

What I did next was pull the raw data for the civilian labor force and apply the participation rate from earlier to it, which gave me a calculated total number of "able-bodied" men and women who might seek employment.  I then calculated the average participlation rate up from October 2002 to the September 2008 (66.1%) and applied that to the "able-bodied" numbers, which gave me an adjusted labor pool.  This adjusted labor pool revealed those who are unemployed but not counted as such.  The number that started low (230k) but as time went on and the participation rate dropped, it reached upwards of 6 million people in 2012.  These people have simply been dropped off of the rolls.

Next I took the raw data for unemployment and adjusted it upwards by adding back those people who had been dropped off the rolls.  Finally I did two calculations - using BLS's original number for labor pool and unemployment, I calculated the unemployment rate.  I did the same with my revised labor pool and unemployment numbers. 

The difference is huge.  The unemployment rate is steady around 12%, and has been since December, 2009.  There are close to 19 million people fully unemployed, a full 6 million more than are being reported.  It's pretty crazy.

Now remember this is just an exercise.  It's important to realize that a portion of those I added back in probably did retire.  But given the steady participation rate of the 20 years (we've been above 66% since 1989) I think that number is small.

Anyway, just having fun here, questioning authority as usual.  Hope you've enjoyed the read.

No comments:

Post a Comment