Each month the BLS compiles unemployment data from a US Census Bureau telephone survey of approximately 60,000 households. As per internal regulation, the Census can only contact prospective survey participants via land line. Therein lies the fatal error: According to the Centers for Disease Control and market reports from the telco industry, more than 1/3 of Americans don't have access to a land line at home, including 50-60% of those age 18-30 and 40-50% of those living below the poverty level. As unemployment rates for the young and the uneducated run nearly twice the national average, there is a high probability that these groups are severely undercounted.

How much are the young and the poor undercounted? Using a fair sample of 50% cell phone respondents, I calculate by more than 2 million. Adding 2 million to the official tally of 14 million unemployed, we get a U-3 rate greater than 10%.

And how much do the 2 million undercounted unemployed add to the consumer spending deficit? If 14 million unemployed contribute to a total of $850 billion in lost annual earnings, 16 million contribute to a total of $910 billion. At the end of next year, when nearly 50% of the unemployed exhaust their benefits, including a disproportionate number of young and uneducated Americans, the total drain on consumer spending will be more than $1.3 trillion.

[snip]

As a side note, the BLS definition of "underemployment" is working part-time hours while desiring full-time work. This definition has nothing to do with skill utilization or pay scale. If the real U-3 unemployment rate is greater than 10%, it would not be out of the question to suspect that the U-6 rate is several points higher than the reported 16%.

The job market continues to be an albatross for investors. Approach this rally with caution.

Bulls Beware: The Real Unemployment Rate Is Much Higher Than BLS Reports
Seth Mason, November 4, 2011

Come January, how will people like me be counted in the White House official talking points? I was self-employed for nine years, had to shut down the business, and therefore was ineligible to apply for UI benefits in the first place. I started working a part-time holiday retail job last week, which will end Christmas Eve. The bottom line: I haven't been counted in the unemployed numbers for the past three years, but I will be counted in the December job creation numbers, and then, once again, I will not be counted in January's unemployment numbers. The final conclusion from White House Office of Economic Indicators will be... that a job was created.

this, of course, is a complete misreading (assuming he ever actually read the actual report) of a recent roper survey/report, on the decrease in movement of people in the US. in fairness to young mr. atkins, he based his entire post on another column, written by someone who also apparently failed to read the source document.

Also, just as a counter to those that think job numbers are always revised down - "the previous two months were revised up to show that 72,000 more jobs added -- the fourth straight month that the government has revised prior months higher"

It does go both ways, and now it seems to be going the right way.

I would also like to point out "Private employers added a net gain of 140,000 jobs in November. But governments shed 20,000 jobs, mostly at the local and state level. Governments at all levels have shed nearly a half-million jobs in the past year."

Many of us have pointed out over the last several years that while jobs are being added, they are not the same kinds of jobs that were lost, and they are not jobs that are compensating people at the same level their previous jobs were.

Today, we get this report, via the NYT:

Even though the Labor Department is expected to report on Friday that employers added more than 100,000 jobs in November, a new study shows just how rare people like Ms. Mowery are. According to the study, to be released Friday by the John J. Heldrich Center for Workforce Development at Rutgers, just 7 percent of those who lost jobs after the financial crisis have returned to or exceeded their previous financial position and maintained their lifestyles.

The vast majority say they have diminished lifestyles, and about 15 percent say the reduction in their incomes has been drastic and will probably be permanent.

[snip]

According to the Rutgers study, those with less education were the most ravaged by job loss during the recession. Even among those who found work, many made much less than before the downturn.

"The news is strikingly bad," said Cliff Zukin, a professor of public policy and political science at Rutgers who compiled the study, which was based on surveys of a random sample of Americans who were unemployed at some point from August 2008 to August 2009. The numbers represent "a tremendous impression of dislocation and pain and wasted talent," he said.

More than two years after the recovery officially began, American employers have reinstated less than a quarter of the jobs lost during the downturn, according to Labor Department figures. Of the 13.1 million people still searching for work, more than 42 percent have been unemployed for six months or longer. About 8.9 million more are working part time because they cannot find full-time work.

[snip]

Even many of those who have managed to find a job are struggling to restore financial stability. "They have had to take pay cuts or benefit cuts or maybe they don't get any vacation," said David Elliot, communications director for USAction, a coalition of grass-roots groups that will release a report on Friday about the experiences of unemployed and underemployed workers.

Not only is 7 percent an astonishingly low number of people who have been able to get themselves back on the same financial footing, but it confirms, I think, that extended high unemployment has significantly depressed wages.

When U3 was at 5.5 in 2003, U6 was almost 11.5.  In December 2008, when U3 was at 7.5, U6 was at 13.5 percent and rising at the exact same rate.

The point is that when trying to determine whether we are improving or falling back, it does not seem to matter whether you are using U3 or U6.  They historically track very closely to each other so using one to determine delta over the other is unnecessary.