Skip to Content
Investing Specialists

Is the Unemployment Rate Really 16.6%?

The U6 takes an already badly lagging economic indicator and puts it even further behind the eight ball.

The formerly obscure measure of unemployment known as U6 has garnered a lot of attention this recession. This data, from the Bureau of Labor Statistics, counts the conventionally unemployed (those able to work, available to work, not working at all, and who have looked for a job in the last four weeks) plus those working part time but preferring full-time work, discouraged job seekers who haven't actively looked for a job recently, and people who want a job but went back to school or took on family obligations, making them unavailable for work.

While conventionally measured unemployment peaked at just over 10%, the U6 unemployment rate peaked at well over 17%. (Both measures were modestly off their highs at 9.7% and 16.6%, respectively, in May.) Some pundits have used this much higher number as an indicator that we are rapidly approaching the 25% unemployment rate of the Great Depression. Others seem to imply that the U6 number is the real unemployment number and the conventional unemployment number is a smoke screen to hide how bad things really are. A recent Wall Street Journal editorial (the full article may not be readable without a subscription) indicated that a lot of the interest in the U6 statistic might be more political than economic, concluding that those on the far right and far left are trying to use the ugly U6 data to score political points.

I strongly disagree. The U6 number just adds a level of complication to the already difficult-to-interpret unemployment indicator. Furthermore, it appears to me that the U6 takes an already badly lagging economic indicator and puts it even further behind the eight ball.

The biggest issue I have with the U6 statistic is that it has been fully compiled only since 1990 and has only been around for two of our most mild recessions. However, we do have data back to 1955 that includes the conventional unemployed and those working part time who would rather be working full time. As the table below illustrates, this part-timers factor captures the bulk of the difference between the conventional unemployment rate and the U6 statistic.

 Unemployment by Category: May 2010
Conventional Unemployment9.7%
Part Time But Wanting Full Time5.7%
Haven't Looked for Work in 4 Weeks0.7%
Couldn't Find Work, So Doing Something Else0.7%
U6 Unemployment Rate16.6%
Source: Bureau of Labor Statistics

Using this data, I have calculated a U6 proxy number (conventional unemployed plus part-timers) back to 1955 and graphed it along with the conventional unemployment number in the chart below:

Note that this U6 substitute number tracks conventional unemployment with a slight lag time. Also note it is highly likely that U6 unemployment in the 1980s was higher than it was during the current recession.

Researcher N Andrew draws a similar conclusion in a research paper that attempted to compute U6 even farther back, including the 1930s.

Though his methodology is different than mine, his conclusion that we are not in uncharted territory is similar. He also posits that U6 unemployment during the 1930s could have been as high as 37%, a long ways from where we were in the most recent recession. He does leave open the possibility that higher debt levels now could make the current situation nearly as difficult as the 1930s. Though I might argue that better government safety nets and far higher asset holdings offset some of that effect.

As an economist, I don't think U6 or U6-like substitutes provide much new information. Also, I am not really comfortable with the concept of someone working 33 hours being counted as unemployed. There is a big difference in spending habits between someone on a temporary one-day-a-week furlough and someone without a job at all. I also believe that it is natural for the number of part-timers seeking full time work to go up during a recession. I know of several instances where a part-time worker immediately sought full-time work after a spouse was laid off. If that person could not find a full-time position, he or she would end up in the "part-time for economic reasons" bucket. In this case, isn't marking both spouses as "unemployed" a bit of double-counting?

Don't get me wrong, this recession was a very bad one with more than 8 million jobs lost. Unemployment, no matter how you measure it, is just a whisker away from the highs of the early 1980s. But we aren't back in the Great Depression or totally uncharted territory, either.

See More Articles by Robert Johnson