The core group of the labour market (almost two-thirds of the working-age population), however, is the 25-54 demographic. Here the OECD says :

There's no trace here of the ten-percent gap conceded by France in the overall figures. The explanation for the higher number of employed in the two Anglo economies has to be found in the marginal groups, 55-64 and 15-24.

The proportion of employed people in the 55-64 age-group differs considerably from one country to the other:

The complicating factor here is retirement. The normal retirement age in France is 60 for both sexes, in the UK 60 for women and 65 for men, in the US at least 65 for both sexes, then, on a rising scale depending on year of birth, 66 or 67. The age group that can be readily compared is therefore 55-59 :

There's a clear under-employment problem with the French over-55s here. Analysis from the Research and Statistics division of the Labour Ministry (DARES) counts 12.7 % of the age-group concerned by two types of measure: early retirement and exemption from job-seeking. The twelve-point gap with the UK and US is there. Early retirement programmes were rolled out mostly in the '80s and '90s then phased out, but they have been replaced by job-seeking exemptions for the over-55s. What was meant to ease the pain of lay-offs by letting older workers (who agreed) go, has become systematic and means French employers rarely hire the over-50s any more (qualification and experience cost more, of course...). Normal retirement at 60, for many years a claim of organised labour and the political left, also plainly reduces the French employment rate calculated on the OECD's standardised 15-64 working-age population.

American attitudes to retirement are different. Not only is the "normal" age of retirement 65-66-67, there are even one million over-75s in employment. The employment rate of the over-65s is 14.6%. It's hardly surprising that the employment rate of the 55-64 group is high (61%).

A BLS study analyses "bridge jobs", often part-time work that comes between the end of the career job and full retirement. These are more common at the top of the earnings scale, and at the bottom. And this of course underscores the ambiguity of late retirement: some may welcome it because they are doing highly-skilled work in which they may have acquired considerable expertise and earn well, while others, because their earnings are low and retirement prospects shaky, may be more or less forced to go on working in order to survive.

At the other end of the labour market, in the 15-24 group, the employment rates are:

The gap is surprising: US and UK rates are double that of France. In all three countries, this demographic weighs 18 to 19 % of the 15-64 population, so the effect of such a wide gap on the total employment rate is not negligible.

(Now for a technical nitpick: the OECD puts together comparative labour statistics that include European countries and the US. It does so on the basis of a standard working-age population from 15-64. As we have seen, that doesn't cater for differences in retirement age. Neither does it cater for differences in youth statistics. Both the US and the UK start their labour statistics at age 16. France, along with a number of other countries, starts at 15. We could go into a digression here about why this is silly (when fifteen-year-olds are legally obliged to be at school), the point here is its statistical effect given that OECD makes no allowance for it.

The population group 15-19 quoted in the OECD stats is taken directly from national labour force surveys; it covers the 16-19 population for the US and the UK, and the 15-19 population for France. Fifteen-year-olds in France are 98 % in school and not in the labour force. Counting in this cohort therefore has no effect on the unemployment rate, which is an intra-labour force ratio. However, by increasing the denominator (working-age population) in the calculation of the employment rate, while not increasing the numerator (employment), it reduces the rate. By a rough reckoning, stripping out the 15-year-old cohort from the French numbers adds three full points to the employment rate.

This does not prevent the rate from remaining low... But one wonders why the OECD doesn't do this as a matter of course?)

Total population in this group (16-24) in the US in 2006 was 37 million, and 45 % of these were not in school. Among those Not In School, the employment rate was 72.7 %.

20.8 m were students, in high school (10.3 m) and college (10.5 m, only 1.6 m of them part-time students). In this group, the larger of the two, the employment rate is 40 %. So, of those in education, 60 % don't have jobs, 40 % do. Of those jobs, 74 % are part-time.

The profile is similar in the UK : around 72 % employment among those not in full-time education, 36 % employment for those who are in full-time education, but with this difference, that only around 37 % of the 16-24 age group are in full-time education. By my calculations (based on data from Eurostat, EU Labour Force Survey), almost all student jobs are part-time.

In France, however, two-thirds of this age group are in education. Only 13 % of those in education have a job, and that includes apprentices for 8 of those 13 %. Student part-time jobs cover only 5 % of those in education.

I calculate that the proportion of student part-time jobs in total 16-24 employment is:

One can argue about the reasons for this difference. There's an historical cultural element in the US and UK that sees students (including secondary-school) as quite naturally having a job on the side, but this is much less true of France. The greater and increasing privatisation of higher education in the US and the UK undoubtedly also exerts greater pressure on students to find money to get through their studies. Is it in order to express a pious wish that full-time students might be just that, full-time students, or is it too, like, passé?

The other point is the rate of participation in education among the 16-24 group: about 66 % for France compared to 55 % for the US and 37 % for the UK. In a society in which it is not axiomatic that a student should have a job, this places a low bound on the 16-24 employment rate. One may argue that higher education here serves to hide youth unemployment. Indeed it's often said that French youth hangs around in the education system for years because no jobs are available. However, the heavy student enrolment in France is in the 16-19 segment, and the OECD's Education at a Glance 2005 gives the percentage of 20-29-year-olds enrolled in education as 20 % for France, 29 % for the UK, 23 % for the US, which would appear to indicate that young people prolong their enrolment more in the latter two countries than in France.

In statistical terms, what influence do student part-time jobs have on the employment rate?

16-24 age group corrected (my calculations) for student part-time:

France 25.8 %

UK 43 %

US 38 %

16-54 (total working-age population) corrected (my calculations) for student part-time:

France 63.6 %

UK 70 %

US 68 %

The ten-point gap in the OECD rates has not entirely disappeared, but the numbers are much closer.

And -- to go back on a point that has been made before on ET -- the 2005 unemployment/population ratio (not the unemployment rate, but the percentage of job-seekers in the total population) for the 15-24 group is:

France 8.6 %

UK 9 %

US 6.2 %

Worth remembering when you're told that practically one in four French youngsters is looking for a job.

Third and final part to follow.