As U.S. jobs disappear at a rapid clip, the official unemployment figure seems understated. While November's 6.7% rate is a full 2% higher than the same time last year, the rate remains well below the 10.8% postwar peak, reached in November 1982. One issue is that the official unemployment number captures only a slice of the total joblessness in the U.S.

To be counted as unemployed in this statistic, a worker must not have a job, be currently available for work, and have actively sought employment within the last four weeks. In other words, a lot of the jobless are left out of the government's tally.

Rajeev Dhawan, director of Georgia State University's Robinson College of Business, says the official unemployment rate is "not a good measure of what is happening in the economy. It's drawn from a sample too small and filled with too many assumptions. Absolute job losses and retail sales give a better idea of what's really happening in the economy."

Fortunately, digging deeper into the labyrinth of the U.S. Bureau of Labor Statistics' (BLS) Web site can offer a more complete, if imperfect, picture of joblessness. Since 1993, the BLS has tracked a category of unemployed called U-6, which captures the total unemployed, plus what the agency calls "marginally attached" workers and those employed part-time "for economic reasons."

For November 2008, that rate was 12.5%, nearly double the official unemployment rate and the highest since the government started tracking this category.

Outside Looking In

Marginally attached workers are those with no job and who aren't hunting for one but who are interested in working -- people who have left the workforce because the employment situation seems so bleak that they've stopped trying. This measure covers anyone who has looked for work in the past 12 months, not just the past four weeks.

In November, 1.9 million workers were marginally attached, up 637,000 from a month prior. This category includes long-term unemployed, such as factory workers who can't find a job paying close to what they'd been earning before. Unemployment rates in construction and extraction jobs such as mining hit 12.1% in November, followed by 9.4% in production jobs. That means the ranks of the marginally attached will increase.

Those employed part-time for economic reasons, who are counted as employed in the official statistic, want and are available for full-time work but have had to settle for a part-time schedule. As of November, the number of workers in this category rose by 621,000. There are now 7.3 million involuntary part-time workers, up 2.8 million over the past 12 months.

Contract workers, sometimes known as freelancers or independent contractors, face a special set of problems when it comes to being counted by the government. First, employers aren't required to report layoffs of contract workers to the government, so when companies say they're cutting their contractor workforce -- as Google (GOOG) did in October -- no one knows by how much.

These job cuts are also not recorded in the official job-cut statistics tracked by the government. In other words, the 533,000 jobs lost in the November count don't include any of the tens of thousands of contract workers being slashed from company payrolls as the recession deepens.

Falling Between the Cracks

Some self-employed workers are incorporated into other BLS statistics, but not all of them are counted. Those traditionally considered self-employed, such as independent real estate agents or accountants, are included in the government's household survey of the unemployed. But those working as long-term freelancers for one particular company without the benefits of being staff members -- often dubbed "permalancers" -- are not.

That means a good portion of this group, which the Government Accountability Office says makes up 10% of the workforce, isn't properly tracked. "We really don't know what is happening with the [contractor employment] numbers," says Sara Horowitz, founder of the Freelancers Union, a 93,000-member organization of contract workers.

Horowitz says the government should develop better measures of contract workers, perhaps by identifying the number of contractor tax filings with the IRS each year. "An increasing part of the economy is driven by this new workforce, but government agencies haven't updated their methods for counting them," she says.

The BLS does capture other pieces of the unemployment puzzle. It breaks out such demographic categories as education levels. As of November the unemployment rate for college graduates increased less than a percentage point, to 3.1%, while the unemployment rate for high school dropouts rose from 7.6% to 10.5%.

The BLS also tracks such categories as age and ethnicity; the unemployment rate in November was 32% for black teenagers, for example. Other data offer state-by-state comparisons of unemployment rates. In the most recent data, which cover the first 10 months of 2008, Rhode Island and Michigan were tied with the highest unemployment rate, at 9.3%, with California next at 8.2%. Though not officially a state, Puerto Rico's rate stands at 12%.

Still, calls for improving the BLS metrics continue. While Horowitz presses for better accounting of contract workers, Georgia State's Dhawan says the surveys need to account for population growth. "Fifty years ago, the [official unemployment] number had some validity," he says. "Now I have little faith in it."

• Asia's Tech CEOs Still Eye Robust Growth
• Plus: Asian IT to Emerge from Financial Crisis Stronger