There was rejoicing here at Eurotrib earlier today when the German and French unemployment numbers were reported, but the numbers can be interpreted differently depending on how you use the statistics--and were done so in the English-language press.

In the German case, a change in the statistical method means that "Unemployment in Germany, Europe's largest economy, jumped in September as people who'd been removed from the Federal Labor Agency's register at the start of the year were reclassified as jobseekers." By the old rules there are fewer unemployed, but by the new rules, more.
http://www.bloomberg.com/apps/news?pid=10000085&sid=aaEyfQkp69ho&refer=europe

In the French case, unemployment fell if you used the national recording method, but stayed the same if you use the International Labor Organization standards.
http://www.businessweek.com/ap/financialnews/D8CUDU9G0.htm?campaign_id=apn_home_down&chan=db

So there is certainly some bending of the rules going on in this whole area.

It seems to me that trying to define exactly what you mean by "unemployment" is pretty tough. A highly paid consultant who is temporarily between contracts is a completely different situation from an unskilled laborer in a weak economy--yet both may be grouped together. And there is the complexity of how to count "working in the home" as employment; surely a mother at home caring for her young children is "working" just as much as the daycare center worker doing the same tasks. This even feeds into the "looking for work" issue, because I know at-home mothers who would work if they could find a job that paid enough to cover daycare costs, but many jobs don't.

It is a statistical nightmare.

NUREMBERG -- Unemployment in Germany jumped this month as people removed from the Federal Labour Agency's register at the start of the year were reclassified as job seekers, the agency said yesterday.

Unemployment numbers, adjusted for seasonal swings, rose 39000 from August to 4,83-million.

Economists had expected a decline of 12000, according to the median of 26 forecasts in a Bloomberg survey. The jobless rate rose to 11,7% from 11,6%. It reached a post-Second World War record of 12% in March.

Labour market figures have been climbing since the start of the year, when German Chancellor Gerhard Schröder's government added welfare recipients to the jobless rolls, cut entitlements and allowed cities and towns to opt out of the federal system.

The changes swelled the number of registered job seekers a total of 483000 in the first quarter, compared with an average monthly increase in unemployment of 10200 last year.

I must save your article and have it handy, when someone in the US has the need to get all upset about those terrible economic problems and the high unemployment rates in Germany to make himself feel good.

Nobody likes to believe me, when I say that someone unemployed in Germany has a very good life compared to someone, who doesn't work in the US and who doesn't appear in their unemployment statistics. I don't like to say this too bluntly on other blogs, because it hurts too many feelings, but it is the truth.

In the US people who work are often homeless and live in shelters or in their cars. In Germany people who don't work have housing, living and educational subsidies and have health insurance. The problem is that Germans expect to have this social security net and become extremely upset easily when the net seem to have some small holes and may leak. The angst is creeping in very fast.

The Americans don't expect much other than not being betrayed too obviously, so they seem to accept or look the other way when people who work still can't make a living with that money and pay their own rent.

Those numbers tell you nothing about people's lives.

It's an honest question - I'd like to have an as definitive as possible answer when that debate crops up.

In the long run, we're all dead. John Maynard Keynes

Bradbury's contention -- that there was more slack in the US labour market than at comparable moments in past cycles, and that the unemployment numbers might be understated by from 1% to 3% -- was taken up by a number of economists and commentators.

Brad DeLong here ,

Paul Krugman here ,

and (useful post with a number of references), New Economist here .

A short refutation of Bradbury's brief was written by Diana Furchtgott-Roth of the Hudson Institute. It can be found (pdf file) here .

Ms Furchtgott-Roth is a well-known free-market pundit whose take-no-prisoners style reminds me somewhat of Margaret Thatcher's. Her refutation seems heavy in breezy assertions and light on relevant substance, whereas Bradbury's paper relies on considerable statistical analysis.

In particular, Furchtgott-Roth states that the teens and women Bradbury deduces are "missing" from the unemployment statistics are in school (teens), at home minding the kids (women), and that this is voluntary, and proves how well the economy is working because "we are prosperous enough to devote more time to education and to our families." (sic). Further, "A decision by couples to live on one paycheck and to let one parent stay home with the children is not a labor market problem." (re-sic).

General Glut picks up the issue and says:

I suggest focusing in on a group of workers who over time are almost always in need of work, who do not wax and wane with educational or retirement opportunities, nor with social trends toward greater workforce participation rates: men age 25-64. ...snip...

As you can clearly see, at the peak of the last two economic cycles the ratio topped out in the 85.5-86.0 range. Currently the level is just 83.2, right where it was three years ago and nowhere close to the levels of the late 1990s.

So: I'm more impressed by Bradbury and supporters than by Furchtgott-Roth. It does seem likely the US official unemployment numbers overstate the dynamism of the US labour market.

End of aside.

People, if you're off to bed, sweet dreams. If you just got out of bed, having read all the above, you'll be going right back.

When locusts move on, they leave nothing behind

I do think we could drive ourselves crazy if we try to adjust these numbers.  As some of the above posts show, professional economists are spending weeks arguing back and forth over some of these concepts, and I doubt if we have the time or the expertise to do that.  A second reason is that if we make our own adjustments (and I think it would really be tough to gain agreement on some of the points), and then want to present that to any outside bodies, I think we would lose credibility by saying something like we used "adjusted' OECD figures as our base.  Isn't that the proverbial Pandora's Box?  I agree it's tempting to do this--the statistic I was most surprised with was the States' assumption that if you work one hour per week, you're employed!  Who ever thought that one up.  But I checked the BLS, and there it was.  But how in the world could we adjust for that?  And I imagine there are others that go the other way, if we really dug into it.

I did have opposite reactions than you to the two points you highlighted on the military and on prison populations.  First, on the military, I was shocked to see that they are not counted as employed, or in the work force.  And perhaps there will be cultural differences on this one.  But we count government employees, except for the military?  I think most Americans would view the military as an honorable profession.  We have prestigious universities focused on the military--the Military, Naval, Air Force and Coast Guard Academies.  These schools are very tough to get into.  Degrees from there are viewed as very honorable whether the person goes onto a career military career or to other professions outside the military.  I guess in the old days of the Draft, I could see an argument that it shouldn't count--but certainly not now, with an all volunteer armed forces.  My review shows there are 1.4 million in the armed forces today, not including reservists.  My SWAG calculation (scientific wild a.. guess) was adjusting the labor force and employed numbers both up by that amount would reduce unemployment by 1/2 of a percent.  And I see your proposed adjustment would add them to the labor force, but count them as unemployed--if i understood you correctly.  Which would mean that the 34th American President went from a lifetime of unemployment to the Oval Office.

On the prison population, the accepted methadologies seem to require that the potential employee be ready to work in one or two weeks.  So there would seem to be a pretty strong argument that they are not in the labour force.  However, I think this statistic is an important one, and should be included somewhere--but probably more in a quality of life area, rather than in unemployment statistics.

Thanks for your great story, and I look forward to continuing dialogue.

I do wonder how I am categorized. I imagine I would be considered a "marginal worker" in both contexts in that I am a graduate student who is effectively unemployed at least 3 months a year. As such, I have to live off my parents and fiance from time to time. This said, being Grad student becomes in many ways a luxury of the fairly well-off and I think accounts for the rather static class nature of academia. Simply put, you have to live on the fringes of the economy for upwards of seven years to become a professor.

Or something close to factual in Bonddad's diary about the middle class.  The whole thing is excellent, but he says this about employment:

According to the Bureau of Labor Services, total nonfarm employment was 132,454,000 in January 2001 and 133,999,000 in August 2005 for a net 4.5 year gain of 1,545,000.  This breaks down to approximately 28,600 jobs/month over the period.  This is important because the economy must create 150,000 jobs/month to deal with natural attrition, people looking for new jobs, people entering the workforce etc....  In other words, Bush's economy has not created enough jobs to deal with the natural expansion of the population or generally economic conditions.  The next question to answer is "why is the unemployment rate so low?"  The unemployment numbers do not count people who have not looked for work in the last 4 weeks.  This is better documented by the labor participation rate which was dropped from 67.2% in January 2001 to 66.2% in August 2005.

I'm not sure, but I think that labor participation rate might have some bearing on things -- do we have other participation rates to compare them to?

Maybe we can eventually make language a complete impediment to understanding. -Hobbes

Regardless of the actual differences in percentage there are plenty of legal and illegal immigrants. Most are coming for economic reasons and are apparently finding jobs.

Somehow the argument about the jobs being unwanted by citizens doesn't seem an adequate explanation. When government or pundit answers defy common sense there is probably more to the issue.

Policies not Politics
---- Daily Landscape