Updating the methodology for estimating entries in France

Since 2014, Insee has been using the annual census surveys to estimate the number of people entering France. One of the main variables used to identify people entering the country, the year of arrival in France for people born abroad, is characterised by partial non-response which was not until now subject of a specific correction. This working paper analyses this partial non-response and proposes an improvement in the quality of this variable of interest, in the form of nearest neighbour imputation. Compared with the methodology used up to now, the method of identifying entries does not change: it is the quality of the data on the date of arrival in France that is improved. This methodological renovation has led to a revision of the series of people entering the country since 2006: their number increases by around 13% each year, while the structure of the population of people entering the country by socio-demographic characteristics or geographical origin changes little. The population residing in France, as well as migration balance, a component of population growth that considers people moving in and out of the country, are not changed by this renovation.