I confess that the most annoying thing about listening to conservative pundits is when they start using economic statistics. Like most of my more liberal ones, they often get them wrong. So reading Jim Geraghty or Ed Morrissey this morning in praise of John Crudele's NY Post column is causing me a little heartburn. Alan Reynolds explains:

The 17.3% figure [the U-6 figure that is all the rage on the right lately --kb], by contrast, starts with those looking for jobs during the past month and adds �all marginally attached workers, plus total employed part time for economic reasons.� That phrase �marginally attached� means people who looked for work at some point during the past year, but not lately. Contrary to press reports, relatively few of the �marginally attached� are those discouraged about job prospects. Adding discouraged workers would only push the unemployment rate up by half a percentage point, to 10.5%. And even that small number of discouraged workers is not simply those who could not find work, but those who simply �think� no work is available, or think they are too young, too old, or that they lack the necessary schooling or training.

The rest of the �marginally attached� don�t even think they can�t find work. Instead, they are not looking for work �for such reasons as school or family responsibilities, ill health, and transportation problems.� To describe such people who are not available for work as �underemployed� (much less unemployed) is an abuse of the language.

As for those �working part-time for economic reasons,� only a fourth say they could only find part-time work. Those who normally work a 9-to-5 schedule (35 hours a week) are counted as working part-time for economic reasons if they miss even one hour �for reasons such as holidays, illness and bad weather.� That isn�t really underemployment, much less �real� unemployment.

These measures have a history going back to 1994, when the Bureau of Labor Statistics developed a way of dealing with an issue found by its former commissioner Julius Shiskin. The whole idea of measuring unemployment is to see if we are efficiently using all our labor resources. But what is "efficient", and what are "all" our labor resources? Reynolds' dissection helps get at what one thinks is unemployed is an opinion about what people would do about their labor-versus-leisure choices in different states of the economy. That means you get inside their heads. And I don't know why you'd pick U-3 or U-6 or U-whatever unless you tell me what you think it means to efficiently use all available labor resources. In 1993 the BLS revised how they assessed those measures by changing the questionnaire (this is why you can't look for U-4 through U-6 before then.) See this article from the Monthly Labor Review in Sept. 1993 on how they changed measuring discouraged workers, in particular at page 24.

A shorter, more recent explanation of the U-# series is here. In a recent working paper from BLS researcher Steven Haugen writes:

These new questions on prior job search and current availability for those who wanted a job but were not currently looking also were used to identify a larger group of persons attached to the job market. Some persons indicate that other impediments, such as transportation problems or child care requirements, keep them from current job search. This larger group of persons (of which discouraged workers are a subset) is referred to as marginally attached workers, and are included in alternative indicator U-5.

The uppermost indicator in the range, U-6, adds in all persons employed part time for economic reasons. Including a group of employed persons is a large conceptual break from the other measures, but since many would agree that these persons are visibly underemployed, those who want to treat the underemployed on an equal basis with the unemployed may find this alternative measure useful.

But as Reynolds has pointed out above, the "part time for economic reasons" is also a state-of-mind concept -- many workers who are on 35 hour schedules consider themselves working full time and not deprived of income. They may not think they are underemployed -- that is left to researchers and pundits to decide. Why? What gives them the insight to know who wants a job and who doesn't, particularly when the Current Population Survey tries to remove subjectivity and takes usually about a 6-minute phone call?

Let me conclude with an anecdote sent by a reader whose large firm which is in "voluntary layoffs."

What surprised me was the attitude of more than one junior engineer also taking the voluntary layoff. In these cases, the junior engineers were viewing the layoff as a sabbatical. They plan to take their severance, then apply for unemployment and planned to mostly use it up before looking for another job.

When a few of us senior engineers heard this story we were amazed. None of us would have ever considered following such a plan during our careers. I guess we are all so work oriented we never would have considered it. I have read cases in the MSM about a "different attitude" toward unemployment during this recession.

Maybe it's different attitudes this time? The Current Population Survey won't tell you a thing about attitudes towards one's job status and neither does U-whatever.

Labels: economics