1. Les mesures du chômage

Unemployment statistics regularly give rise to controversy in the public arena, with the charge that they do not reflect reality. This recurrent allegation is due, in the first instance, to the fact that there are a number of competing methods for measuring unemployment. In particular, it is important to distinguish between the monthly figures for job-seekers (demandeurs d?emploi), published by the French Ministry of Labour, and those for the unemployed, produced by the International Labour Organization (ILO), published by the French National Statistical Office (Insee), and used by them to calculate unemployment rates. Thus, on the one hand we have an administrative calculation, widely-reported in the media, but which may suffer in terms of accuracy from changes in the way unemployment is calculated; and on the other, an ?objective? statistical measure, widely-respected by statisticians in the field for its internationally-accepted method of calculation and because trends are not affected by administrative changes in the way unemployment is managed. Self-assessment by job-seekers in the French census and other surveys represent a third source of data. For several decades now, statisticians have been repeating that these three methods of calculating unemployment are distinct; that they do not have the same objectives or advantages; that they do not, at the end of the day, measure the same thing. But this has not stopped confusion and controversy from persisting.