The data from the ACS of year N are delivered to internal users in the geography of collection, i.e. on 01/01/N-1.

Since 2004, France has revamped its census operation, which is now based on a five-year cycle: each year, only part of the population is counted during an annual census survey (EAR). A annual census survey (ACS) is therefore based on the aggregation of data from five EARs.

The revamped census is therefore not initially designed to produce results for a single year of data collection. The weights calculated up to the 2018 EAR only partially correct for imbalances in the rotation groups.

In order to improve the correction of these imbalances, the “Census Methods and Processing” division

has implemented a new method for calculating

weights. This new method has been

used to calculate weightings since

the 2019 EAR. Although the data are now more robust,

caution should nevertheless be exercised when using data from a single EAR, particularly when exploiting sub-national results.

Furthermore, rate estimators should be preferred over level estimators.