The French have higher productivity than just about anybody but Belgium and the Netherlands. Higher than the U.S. higher than Germany, much higher than Great Britain, much higher than Japan, a bit higher than Ireland. "Poor productivity"? Obviously headline writers don't read OECD statistical tables or perhaps even the article they're writing headlines for.
Productivity, for those who haven't heard, is computed as output over hours worked. More output in fewer hours equals higher productivity. I shouldn't be too hard on a poor headline writer though, considering what the OECD itself is capable of. Below is a short piece I was working on in response to a sighting I came across in the OECD Employment Outlook:
Thinking inside and outside the Box
Box 1.1 on page 28 of the OECD 2004 Employment Outlook (http://www1.oecd.org/media/econsurv/EMO_e_04.pdf) begins as follows:
“The negative cross-country correlation between the employment-population ratio and average annual hours per worker probably does not reflect a demand-side trade-off, in which a more or less fixed volume of work must be shared across the adult population (...). Rather, the response of labour supply to long-run improvements in productivity and living standards appears to differ along the intensive and extensive margins.”
We know that correlation doesn’t imply causation. But what’s this about?
The key to unraveling Box 1.1’s circumlocutions may well be in the elided phrase in parentheses. If you know the Sandwichman, you can probably guess what the three dots censor. Whatever you do, the Box seems to insist, please don’t assume the negative correlation between average annual hours per worker and the employment-population ratio implies causation from reduced hours to increased employment. Assume instead that such a causal relationship is untenable. In this case, perhaps, correlation may even imply non-causation. Quick, find another explanation -- even if it isn’t really an /other/ explanation.
In spite of Box 1.1’s “probably not” and “rather”, productivity smuggles annual average hours per worker back into the equation as a divisor. Which came first, the chicken or the egg? -- the output or the hours? Wouldn’t it be more prudent to say that differences in productivity represent an indissoluble combination of differences in output and hours of work such that elevating either output or hours to the status of prime mover is presumptuous? Which is to say that if total employment indeed responds positively to improvements in productivity, one /cannot rule out/ the possibility that a reduction in annual average hours per worker /contributes/ to the increase in employment.
Now it’s time to restore the suppressed phrase to the parentheses: (the so-called, “lump of labour fallacy”). The novelty of the lump of labour fallacy claim is that it assigns the weakest argument to advocates of shorter working time, “they assume that there is only a fixed amount of work to be done.” And voila, all one has to do is show that the amount of work is not fixed and the case for shorter work time is demolished! That is what is known as a straw-man argument -- substitute a weaker argument for your opponent’s argument and defeat that weaker argument.
Box 1.1 has gone a step further. After defeating (or at least dismissing as not probable) the weaker argument, it unwittingly presents the frame of a stronger argument. Of course it mistakes this stronger argument for a further refutation of the weaker argument rather than as a vindication of the prospects for creating employment through the reduction of working time.