Annual estimates: data are revised between the provisional estimate and the final estimate due to

1) the change in source used for salaried employment (BRC or DSNs processed by Acoss, the Epure chain and Dares in the provisional estimates, against DADS et DSN in the final estimates)

2) the change in treatments applied to source data

3) new data available on self-employment.

Quarterly estimates: data may be revised from one quarter to the next, to correct a processing error or other anomaly which was not identified in the previous quarter.

Revisions between publications (quarterly or annual) are systematically presented and commented on. The most important ones are explained in detail in the methodological sheet attached to each publication.