Why the real unemployment rate is far higher than the official one
Dec 13th 2009 12:00PM
Updated Dec 14th 2009 1:47PM
First, any unemployment number is in effect a snapshot of a moving target: No matter how rigorous the collection process, it's not possible to survey every household and every employer (not to mention the millions of self-employed).
Second, the methods used to collect and adjust the data can skew the results dramatically. For instance, though the headlines announced that "only" 11,000 Americans lost their jobs last month, other data (both government and private) suggests the real number was much higher.
Third, the human bias toward presenting bad news as "less bad" whenever possible is always in play when it comes to official statistics and reports. (Bearers of bad news tend to have short careers.)
While that bias is understandable, glossing over unpopular or undesirable facts often leads to pernicious consequences. If you were facing a dread disease, would you want your doctor to massage the data so that it sounded "less bad," essentially offering you false reassurances, or would you want the truth? Without an honest diagnosis, no patient can hope to improve.
As a result, it is essential to ask if the methods of calculating unemployment yield politically expedient results rather than honest ones, not by fabricating numbers but by burying or obfuscating inconvenient data.
Deciding Who Counts as Unemployed ... and Who Doesn't
The Bureau of Labor Statistics derives its "headline" unemployment number (identified as U-3 in BLS charts) by tabulating data collected from the state unemployment offices. These offices keep tabs on laid-off workers collecting unemployment insurance benefits.
So far so good. But once an unemployed person exhausts their benefits (or extended benefits, "Emergency Unemployment Compensation") and stops reporting their weekly job search efforts to the state office, then they fall off the roster of unemployed people and are instead counted as "marginally attached to the labor force." This category also includes so-called "discouraged" job seekers and those who have formally stopped looking for work for reasons other than discouragement.
There are 2.3 million "marginally attached" unemployed by the BLS count.
Various other unemployed people don't make it into the official count. For example, those people who are hired but let go before they reach 14 weeks of employment -- the period needed to qualify for unemployment insurance -- won't be counted under "initial jobless claims" even though they lost their jobs.
Clearly, undercounting the unemployed lowers the official unemployment rate.
You Don't Really Want to Work, Do You?
Another way the number can be skewed is to drop people from the workforce. If someone loses their job, they're unemployed; but if they're no longer counted as part of the labor force, then they are, in effect, no longer unemployed. They simply vanish into the statistical ether.
According to the BLS, 291,000 citizens left the labor force in November. This means 291,000 people who would have been counted as unemployed evaporated from the official count last month. Skeptics look at the numbers and note that if removing 291,000 people from the labor force left "only" 11,000 additional unemployed, then 302,000 people actually lost jobs.
Critics also suspect the bureau's "birth-death" model of estimating new job creation is essentially smoke and mirrors. Since it is not possible to identify every new small business in the $14 trillion U.S. economy, the BLS performs a black box calculation (that is, one which is not transparently calculated) to estimate how many new jobs were created and lost by the birth and death of small enterprises.
The BLS model may well have produced reasonable approximations in prosperous times, but even the mildly skeptical question the model's accuracy in times of deep recession. The BLS itself notes that "The most significant potential drawback to this or any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships and therefore is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend."
In other words, the model works until it doesn't. According to the BLS model, the economy added 1.18 million new jobs from February through November -- phantom jobs which reduced the number of unemployed (officially 15.375 million).
Asking Real People for Real Numbers ... And Still Coming Up Short
To its credit, the BLS attempts to overcome these deficiencies by conducting a Household Survey which captures other data, such as the 9.3 million workers currently involuntarily working part-time (what the BLS calls "working part-time for economic reasons") who want full-time jobs.
The BLS combines these involuntary part-time workers, discouraged job seekers and those drawing unemployment insurance benefits into a more comprehensive measure of unemployment/underemployment, U-6, which currently stands at 17.2% -- 7.2 percentage points higher the "headline" rate of 10%.
Despite these efforts to include the broadest measure of unemployed, commentators note that official statistics still fail to account for the millions of self-employed citizens who may have seen their incomes slashed to near-zero. (Unless they pay unemployment insurance, the self-employed do not qualify for unemployment benefits.)
Thus, for example, real estate agents who haven't sold a house in months or consultants who have lost their primary contracts are still counted as "employed" in the household survey.
It falls to politicians and the media to interpret all this complex and occasionally contradictory data, and that's where the spin revs up. To take but one example, few of the commentators seeking to present a "less bad" point of view mention that the long-term unemployment number -- which measures how many job seekers are still without work after 27 weeks or longer -- has been rising. According to the BLS, the number of long-term unemployed rose by 293,000 to 5.9 million. These long-term unemployed now constitute 38% of the unemployed -- a percentage which many see as evidence that the U.S. is facing serious structural unemployment.
According to the U.S. Department of Labor, "States reported 4,178,780 persons claiming EUC [Emergency Unemployment Compensation] benefits for the week ending Nov. 21, an increase of 327,729 from the prior week. There were 729,256 claimants in the comparable week in 2008."
Skeptics find little comfort in a "headline" job-loss number which they recognize has been artificially lowered to 11,000 by dubious adjustments. Instead, they see the increase of 327,000 in long-term unemployed as the real data which is all too often glossed over or ignored in politically convenient sound bites and headlines.
Charles Hugh Smith writes the Of Two Minds blog and is the author of eight books, most recently Survival+: Structuring Prosperity for Yourself and the Nation.