Laurent Toulemon, Institut National d'Études Démographiques (INED)
Since 2004, the French institute of statistics has organized a new type of population census, based on annual census surveys. The aim of the paper is twofold. First, it presents, for the first time, an estimate of double counts within this new census using data from the Demographic panel. This panel is a subsample whose size has been set at 4% of the total population, and which includes all census forms, including cases where two forms are completed in the same year for the same individual. Second, by comparing, on a macro basis, the proportion of people with multiple residence and the double counts, the paper aims to estimate the risk of double counts in cases of multiple residence, for specific subgroups of the population. This is done using information on people with more than one usual residence drawn from another dataset constructed by pooling several household surveys. These surveys use the same household grid, which includes questions on the existence of another usual residence. Surveys show that multiple residence is more frequent among young adults, when they are still living in the parental home but also in another dwelling (especially for their studies). Multiple residence is also frequent among children with separated parents, adults working away from their family home, and older adults around retirement age who spend part of the year in their secondary home (which is becoming their usual residence). Preliminary results show that some 40% of people with more than one multiple residence are counted twice in the census. Using age, family situation and occupational status as our main categories, we can accurately estimate the probability of double counting in cases of multiple residence among the French population, and the family situations associated with these double counts.
Presented in Session 69. Data sources and methods quality