Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.

The Unemployment Measure

by Laurent GUERBY Sun Sep 16th, 2007 at 05:03:19 PM EST

Baker & al paper "Are Protective Labor Market Institutions at the Root of Unemployment? A Critical Review of the Evidence" page 5 cites OECD unemployment rate by sex and age: for example for male 25-54 the unemployment rate is 4.6% in the USA and 7.4% in France. The paper then continue using the employment measure without any further questioning of what the unemployment number means.

If one digs a bit more into OECD data, for the exact same sex/age group male 25-54 the employment to population ratio is 86.3% in the USA and ... 86.7% in France.

So the unemployment rate for this population is shown to be 60% HIGHER in France than the USA, yet MORE members of the population in question do have a job in France than in the USA (with similar part-time job rates).

This puzzle was shown to Baker in january 2007 and inserted into french and english wikipedia - where it is still present after many edits.

Raymond Torres, Head of the Employment Analysis and Policy Division, OECD, said to Le Monde (french newspaper) in may 2007 that "unemployment was a less and less relevant indicator to judge the job market efficiency".

Yet, as shown in the Baker paper, even economists who challenge the orthodoxy do not mention this piece of data when doing cross country so called "unemployment" surveys.


Are there authors who have a beginning of explanation of the puzzle I mention here? Why are economists still using the unemployment measure for anything?

If unemployment is used to measure some form of pressure in the job market, what are the transition rates for unemployed vs inactive to holding a job per unit of time?

Notes:

Display:
See also:
Chiffres : ce qui compte c'est l'emploi, pas le chômage | Telos by Bernard Brunhes 04 Mai 2007

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet
by Melanchthon on Sun Sep 16th, 2007 at 05:51:55 PM EST
Unemployment is low in Anglo-Saxon countries, and high in continental Europe, therefore demonstrating the superiority of capitalism overState action.

Nothing can go through the fog of the narrative.

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

by Jerome a Paris (etg@eurotrib.com) on Sun Sep 16th, 2007 at 05:58:41 PM EST
of our ability to create a new narrative against that.

Employment?
Quality of life?
Living standards?

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

by Jerome a Paris (etg@eurotrib.com) on Sun Sep 16th, 2007 at 05:59:49 PM EST
[ Parent ]
This summer I brought this up to my Macroeconomics teacher. Indeed he was a little surprised, since the necessary conclusion is the irrelevance of unemployment rates to compare labor markets across countries. Of course the next conclusion is that Unemployment rates tell little of labor markets within one country... but that's another matter.

Above Jerome asks what kind of narrative we can use to respond. I just realized that we use the U rate to talk about the number of people that aren't living like 'the rest of us' and not to talk about the number of people that can't support themselves, or are left out of the market. This prevents us from looking into this other matter which is what being employed, being on the normal side of things, means in practise.

So while people always focus on this number that's soooo far (60% ! non mais vous vous rendez compte?) above the US' rate, complain, bitch and moan and succomb to neolib reforms, they assume that unemployment conditions haven't changed.

I guess i'm just rallying to ET's outcry of "NUMBERS LIE". Alas, it's hard to contribute here. To really do so, I should propose talking points... last quarter median monthly income, benefits included.... euh...

Cheers

Rien n'est gratuit en ce bas monde. Tout s'expie, le bien comme le mal, se paie tot ou tard. Le bien c'est beaucoup plus cher, forcement. Celine

by UnEstranAvecVueSurMer (holopherne ahem gmail) on Sun Sep 16th, 2007 at 11:11:15 PM EST
All things being equal, unemployment rates tell us something about the short-term performance of the labour market. That's what it was designed to do and that's all it's good for. Comparing it across different environments, geographic, economic or political, is just plain silly and any economist who doesn't know that is either incompetent or dishonest.

As Laurent will point out, you can't even rely on it if there are administrative changes in practice that alter the behaviour of the labour market.

by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 04:48:43 AM EST
[ Parent ]
Unemployment is only describing the government performance in pushing the unemployed into inactive status.

Nothing else.

That's what thethe data point cited in my diary tells us.

Male 25-54 is a very special category since they will constitue the biggest part of the work force and they have no reason not to be working whatever the country/culture, except for real reasons:

  • really sick/disabled and so unable to work
  • in jail / legally prevented for working
  • rich enough no need for work/money (but those will likely report as self employed investors)
  • "normal" transition from job to job
  • looking hard but really no work at all

This is economists work to sort this out, but this work is not done at all and as shown by the paper references millions are spent on economists looking for totally useless ideological stuff.
by Laurent GUERBY on Mon Sep 17th, 2007 at 06:50:53 AM EST
[ Parent ]
What a  set of assumptions you manage to pack into a small space.You left out retraining or education from the things middle aged men are allowed do, for a start.

The point of the unemployment numbers is that they're easy and cheap to collect and are good short-term proxies to changes in conditions if no-one is fucking with the measurements.

"Labour market" isn't an insane concept either: ignoring the ramifications of it is. In fact, without the concept, how do you distinguish between people who have chosen not to work and those who want to work but can't?

by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 07:07:11 AM EST
[ Parent ]
Measuring unemployment means having people going out and asking socially sensitive questions about the scale of their "want" for work and sorting the results. If you join together various country statistical systems I'd bet you'll end up with more than fifty possible answers (OECD does a lot of work to unify the mess).

That's hard to very hard to impossible to make anything out of it.

Measuring employment is mostly just measuring payroll  paid every week/month and various taxes for non wage workers, good job for computers, mostly impartial.

So sorry but "easy and cheap" is not for unemployment but for employment.

Retraining is somewhat normal activity between jobs, I don't know if it is statistically significant amongst other source of between job activity (it should be but well ...). My engineering school has many people going from technician to engineer in back to school programs, easy to spot in the class because of age differences :).

by Laurent GUERBY on Mon Sep 17th, 2007 at 07:51:58 AM EST
[ Parent ]
But the "employment number" is only useful by itself if you're willing to be just as thick-headed as the people relying on the unemployment number.

Talking about unemployment and employment is complicated and requires that you sit down and look at a range of indicators and qualitative factors before you can do any comparisons. It doesn't matter which gross over-simplification you indulge yourself with, it's still nonsense.

by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 08:04:49 AM EST
[ Parent ]
You're right when you say that, whatever the indicator we chose, we must sit down and discuss it. However, as afew says below, not all indicators have the same relevance: some indicators are less reliable and more likely to be biased than others, and this is the case of unemployment rate.

And the fact that it is used without being discussed by economists, journalists and politicians to promote the hardest laissez-faire economy makes it an important issue.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet

by Melanchthon on Mon Sep 17th, 2007 at 08:31:09 AM EST
[ Parent ]
Change to a different single statistic and they'll manipulate that instead.
by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 08:32:00 AM EST
[ Parent ]
It is true that any indicator can be spun. But  employment rate, being simpler, is less easy to spin than unemployment rate.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet
by Melanchthon on Mon Sep 17th, 2007 at 08:40:39 AM EST
[ Parent ]
But it's at least as meaningless: is a 95% employment rate a good thing or a bad thing?
by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 08:57:52 AM EST
[ Parent ]
It is a different question. The quality and reliability of the indicator (and of the measure) is critical if it plays a role in the decision-making process. What the relevant value is depends on the policy you want to promote.

I cannot say if driving at 50 mph is better than driving at 100 mph, but I know it's better to use the car wheel's rpm (with an accurate and reliable speedometer) than to estimate the car velocity by measuring the angular speed of the roadside trees...

.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet

by Melanchthon on Mon Sep 17th, 2007 at 09:32:52 AM EST
[ Parent ]
95% employment rate is probably bad for 15-24 year old, a point I make each time someone compare 15-64 employment rates between countries :).

For male 25-54 95% is reached by very few countries and usually not sustained.

by Laurent GUERBY on Mon Sep 17th, 2007 at 01:08:03 PM EST
[ Parent ]
It's not really about saying the 25-54 male employment number is useful by itself but the employment number is damn useful in showing the unemployment number - relied upon blindly by 100% of the employment debate party - is highly misleading and should be avoided.

See the immediate reaction of UnEstranAvecVueSurMer and his/her Macroeconomics teacher :).

by Laurent GUERBY on Mon Sep 17th, 2007 at 02:17:17 PM EST
[ Parent ]
I agree. However, I think the range should extend to 60 because in most countries, the legal age for retirement is 60 or above.  

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet
by Melanchthon on Mon Sep 17th, 2007 at 07:13:38 AM EST
[ Parent ]
That's a possible extension, I used 25-54 because statistics are usually available only in this range.

However, not everyone is working on a chair :). There are some activities requiring good health and strength/endurance but I don't know what share these jobs have nowadays.

by Laurent GUERBY on Mon Sep 17th, 2007 at 07:37:11 AM EST
[ Parent ]
not everyone is working on a chair :). There are some activities requiring good health and strength/endurance

Having worked for years as a mechanic in the industry, most of the time working in shifts, I know that damn well.

However, it is not relevant: the vast majority of the people who work under difficult working conditions are employed in the private sector, where the legal retirement age is 60, whereas some public sector employees working "on a chair", as you put it, can retire at 55. Anyhow, the number of workers who can legally retire before 60 is limited.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet

by Melanchthon on Mon Sep 17th, 2007 at 08:02:05 AM EST
[ Parent ]
"Preretraite" is common is the french industry (since it is actively promoted by french government...) and is likely the cause of the low employment rate, my grandfather was a factory worker got it before 60 IIRC.

I'd love to get more accurate statistics about 55-59.

by Laurent GUERBY on Mon Sep 17th, 2007 at 01:11:54 PM EST
[ Parent ]
Early retirement has been widely used by (mostly big) companies because it was the easiest way to get rid of workers without troubles (the retirement system (i.e. other workers) sharing the burden of the costs). Until recently, the governments encouraged this kind of practice to avoid social unrest.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet
by Melanchthon on Mon Sep 17th, 2007 at 01:28:13 PM EST
[ Parent ]
I've just come across an interesting paper looking at the impact of disability and gender on employment of disabled people. It breaks down the data into those working (and in what sector), those who are unemployed and those who are inactive.  Haven't had chance to read properly or absorb but will comment more once I have.

Across the UK the average of 40% disabled people ar ein employment. In Wales, this drops to around 26%, the lowest in the UK.

But anyway, the point I want to make here is that you missed out discrimination.  Plenty of disabled people are able to work but can't find jobs because they are discriminated against in the recruitment process.

by In Wales (inwales aaat eurotrib.com) on Mon Sep 17th, 2007 at 07:25:51 AM EST
[ Parent ]
For disabled I know a bit about the french situation since my sister work in helping disabled people have a normal work life but classification of ability is really hard to do.

However disparities between areas and countries for disability as you mention should be scrutinized closely.

Discrimination is of course included in my last category.

by Laurent GUERBY on Mon Sep 17th, 2007 at 07:40:11 AM EST
[ Parent ]
The same argument of discrimiantion preventing employment can extend to ethnic minority groups, possibly women as well, but as you say, classifying this is overly complicated.

Although with disability - severe health related disability would prevent working or disabilities which require say 24 hour supervision and care. But still many disabled people would be assumed to be unable to work when actually this may not be the case - as I'm sure your sister has found.

It would be interesting to get a comparison of the rates of employment across countries for disabled people because it could say a lot about access to employment in different regions.

by In Wales (inwales aaat eurotrib.com) on Mon Sep 17th, 2007 at 07:47:55 AM EST
[ Parent ]
Numbers don't lie: people compiling and using numbers lie.
by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 04:51:08 AM EST
[ Parent ]
I guess  Numbers are only worth how they are used... so the circle is closed.

I wasn't really addressing what the U rate can tell us, but more what we shy away from by using it all the time.

With regards to your comment above about being a proxy for short term changes, I forget what the exact numbers are, but approximately 3 million people drop out of the labor force every month in the US, and then being categorized either as unemployed or inactive. 1.5 million people go from inactive to employed in the us every month.

Rien n'est gratuit en ce bas monde. Tout s'expie, le bien comme le mal, se paie tot ou tard. Le bien c'est beaucoup plus cher, forcement. Celine

by UnEstranAvecVueSurMer (holopherne ahem gmail) on Mon Sep 17th, 2007 at 09:41:30 AM EST
[ Parent ]
France IIRC has a noticeable rate of job creation/destruction, something like 10 000 jobs per day.
by Laurent GUERBY on Mon Sep 17th, 2007 at 01:20:01 PM EST
[ Parent ]
See this diary from last year, quoting Blanchard:


The labor market is characterized by large flows--high rates of separations from firms, and high rates of hires by firms. In France for example, 1.5% of all jobs are destroyed each month and roughly as many are created-- interestingly, this is about the same percentage as in the United States. As there are many reasons other than job destruction why a worker may separate from a firm, the flows of workers are typically much higher. In France, they are of the order of 4% per month (Pierre Cahuc and André Zylberberg 2004).


In the long run, we're all dead. John Maynard Keynes
by Jerome a Paris (etg@eurotrib.com) on Tue Sep 18th, 2007 at 04:59:08 AM EST
[ Parent ]
A big more on end of jobs statistics in France here:

http://ew-econ.typepad.fr/mon_weblog/2007/03/questions_ouver.html#comment-66232414


A Laurent: certaines des réponses se trouvent dans ces deux notes assez complètes de la DARES.

http://www.travail.gouv.fr/IMG/pdf/publication_pips_200307_n-28-2_nouveaux-usage-licenciement-motif- personnel.pdf

et

http://www.travail.gouv.fr/IMG/pdf/2006.03-11.1.pdf


Merci pour les liens.

Donc la réponse (tableau 2 page 6 du premier PDF en recalculant hors retraite etc sur la premiere colonne) est 71% de demission, 20% de licenciement raison personnelle et 9% de licenciement economique. Les papiers remarquement que les licenciement perso+eco se somment a peu pres a constante et qu'on observe une substitution.

La contestation concerne donc 20% de 20% plus epsilon des economiques soit
4% des fins de CDI.

En cherchant via google j'ai trouvé ce PDF :

http://www.ca-lyon.justice.fr/fichiers/CONSIGNY/Bicentenaire/Conference_JEAMMAUD.pdf

Page 6 il est dit que 75% des requetes des salariés sont gagnantes (au moins partiellement) aux prud'hommes (est-ce que des économistes se sont penché sur cet indicateur ?).

En appliquant ce chiffre il reste 1% des fins de CDI qui se terminent en recours jugés abusif des salariés.

Pour couronner le tout, les cadres sont sans doute bien surrepresentes dans ce 1%, ils ont les moyens de jouer.

Bref, a la louche les incitations discutées ici c'est plutot du blah blah.

Si on veut améliorer le bien être, il faut plus regarder comment reduire les délais au prud'hommes, ramener les resultats a 50%-50% et sans doute le plus gros impact en bien etre est a regarder du coté des démissions.

On peut en effet sans doute en offrant un petit avantage au salarié lors d'une demission (actuellement il perds absolument tout) pour reduire la part des licenciements avec globalement du gagnant-gagnant grace a la baisse de l'insecurité juridique via les recours.

Il y a des etudes sur ce vrai sujet ?


Les salariés gagnent car les entreprises preferent jouer les prud'hommes en depassant largement les bornes plutot que de negocier des clauses acceptables de licenciement, la division par 4 est donc justifiée, les employeurs sont juste tres mauvais payeurs, un comportement a ne pas encourager.

Ensuite les chiffres des papiers (je ne sais pas pourquoi) ont l'air d'exclure les entreprises de moins de 10 salariés (quelle est leur part dans l'emploi et les licenciements+demission ?). Donc raisonner sans chiffres sur une entreprise de 5 salariés me parait risqué ...

Le papier que je cite donne des chiffres autour de 200 000 recours au prud'homme par an, soit 15-20% du total des recours a la justice (> 1 millions d'affaires) et moins que le nombre de saisies bancaires et la moitie du nombre de gardes a vues (plus de 500 000 en 2006, avec les humiliations qui vont avec).

Le nombre de recours aux prud'homme a baissé de 7% en 10 ans :

http://www.journaldunet.com/management/dossiers/0604132prudhommes/index.shtml

Et cela dans un contexte d'augmentation global du contentieux par ailleurs, la part prud'hommale doit etre bien en baisse.

Donc non ce n'est pas un gros chiffre pour commencer.

Et la ou je ne suis pas d'accord c'est le levier, la premiere fois que les demissions sont évoquées c'est dans mon commentaire.

Vu leur importance relative (20 fois le nombre de recours et 70 fois le nombre de recours abusifs par un salarié), une incitation bien plus faible est necessaire a leur niveau plutot que sur le licenciement pour avoir un impact et le plus important c'est gagnant-gagnant et pas juste perdant pour le salarié en argent et en sentiment de justice comme le reste des propositions avancées ici.

Un indemnité (au hasard) d'un mois de salaire (utilisable une fois maximum par periode de deux ans) pour une demission avec une partie (50% ?) payée directement par l'employeur et l'autre par l'assurance chomage aurait sans doute aussi des effets positifs dans l'economie dans son ensemble (sans parler des relations employeurs-employés).

Rédigé par: Laurent GUERBY | le 12 avril 2007 à 23:27

by Laurent GUERBY on Tue Sep 18th, 2007 at 02:43:26 PM EST
[ Parent ]
MSM/economists narrative works because most "normal" people reasonably think that unemployment follows the "normal/obvious" definition of counting those without job out of the total population. Even economists sometimes do think so.

But there is the completely stupid notion of "labour force" which is total fantasy and worse than useless in the middle.

The OECD data point as you say will immediately terminate any discussion of using the unemployment number for anything (except joking about economist = stupid).

Yet, no economist has adressed it.

It says a lot.

by Laurent GUERBY on Mon Sep 17th, 2007 at 06:54:57 AM EST
[ Parent ]
Even if numbers can mean next to nothing and we accept the idea that the concept of the Labour Force is useless, there remains the question of structural unemployment and the ability of a given economy to employ those whose jobs have disappeared because of changes in the labor market. I am thinking of coal miners in France, autoworkers in the US, etc. (analysts in NY? ^^)

This kind of unemployment has lasting social and economic consequences in entire regions such as reactionary voting, low education standards, mass emigration, etc. This could also be applied to ethnic groups which are consistently underemployed and poor (blacks in the US).

For those people, the unemployment number can tell us something, if coupled with average duration of unemployment.

Rien n'est gratuit en ce bas monde. Tout s'expie, le bien comme le mal, se paie tot ou tard. Le bien c'est beaucoup plus cher, forcement. Celine

by UnEstranAvecVueSurMer (holopherne ahem gmail) on Mon Sep 17th, 2007 at 09:58:46 AM EST
[ Parent ]
The numbers are not meaningless nor do we need to throw away the concept of a labour force - though that's always something fluid. The point really is that the picture is complicated and attempting to concentrate on any single figure is simply deceptive. That's why the the US BLS collect a range of statistics rather than a single figure as do many of the other national statistical bureaus. That these are generally ignored by the pop economists is just another symptom of the superficiality with which we treat economic policy.
by Colman (colman at eurotrib.com) on Mon Sep 17th, 2007 at 10:06:50 AM EST
[ Parent ]
Agreed.

Maybe there is some kind of "plague of the indicators". We have a few numbers that we use to make sure the plane is still flying, when really we might as well be in free fall, to use a convoluted metaphor.

I rephrase what i responded to Laurent: that we can use those tools, that there is no reason to dump them away, but that they should be applied to specific segments of population. Using them to describe whole countries gives us this false sense of remoteness that prevents us from looking in the details -- labor condition and living standards -- and facilitate the propagation of preconceived messages.

Rien n'est gratuit en ce bas monde. Tout s'expie, le bien comme le mal, se paie tot ou tard. Le bien c'est beaucoup plus cher, forcement. Celine

by UnEstranAvecVueSurMer (holopherne ahem gmail) on Mon Sep 17th, 2007 at 10:44:55 AM EST
[ Parent ]
It may be that at one time in the past, when employment was more stable, the Unemployment Rate might have borne some useful information - in a time series for a single country, not as an indicator used to compare economic "performance" between countries.

Nowadays, when the notion of "labour force" seems less cogent, when people move in and out of short-term jobs and large numbers of people work part-time and even very short part-time, when the definition of the "inactive" category is fuzzy, the UR is no longer a useful metric. (The Employment Rate would be more useful.)

In my diary Soundbite Statistics: the Unemployment Rate (1), I examined the employment situation of the 25-54 age group in France and the UK for the year Laurent cites above (2004). Conclusion: the difference in the UR is due to the difference in definition of "inactivity", particularly under the heading "long-term sickness/disablement".

by afew (afew(a in a circle)eurotrib_dot_com) on Mon Sep 17th, 2007 at 07:58:29 AM EST
And it is critical to question it because it is a key feature of the dominant economic doxa (and therefore of the media doxology).

As we had discussed before, I think we should look at the full-time equivalent employment rate in order to eliminate the effect of different part-time employment figures between countries.

"Dieu se rit des hommes qui se plaignent des conséquences alors qu'ils en chérissent les causes" Jacques-Bénigne Bossuet

by Melanchthon on Mon Sep 17th, 2007 at 08:19:30 AM EST
[ Parent ]
Via Vonric

http://vonric.blogexpat.com/blog/propositions/2007/09/16/le-chomage-au-royaume-uni-6-fois-plus-lev-q ue-les-chiffres-officiels

http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2007/08/16/nwork116.xml


Nearly 10 million adults in Britain are currently out of work, one in four of the working population, the Government admitted yesterday.

Philip Hammond, the shadow chief secretary to the treasury
Philip Hammond: 'Many of these would want to work'

Official figures showed there are 1.65 million people who are unemployed, with a further 7.9 million defined as "economically inactive".
[...]


by Laurent GUERBY on Tue Sep 18th, 2007 at 04:53:25 PM EST


Display:
Go to: [ European Tribune Homepage : Top of page : Top of comments ]