The new census housing questionnaire and the redesign of the Households and Families’ analysis in 2018: What contributions for the study of family structures?

Since the 2018 census redesign, the housing questionnaire of the census describes the two to two relationships between housing dwellers and includes questions on multi-residence. The redesign has also concerned the so-called "Households and Families’ analysis" (AMF, for “Analyse Ménages Familles” in French), i.e. the statistical processing of information on the relationships between people living in housing to define families.

As the census has become a multi-mode survey (paper and internet), shortly before the 2018 redesign, it is necessary to take into account the differences between the two modes of collection. For example, declaring that they live as a couple is much more frequent among internet respondents, whereas the question is asked in exactly the same way on the internet and on paper, and this greater propensity to live as a couple cannot be explained by the observed differences in the characteristics of the people responding on the internet (younger and more qualified in particular).

With regard to family types, the collection method is also important: the partial non-response on the links between inhabitants of the dwelling is frequent on paper and rarer on the internet. This makes the imputations made by the AMF particularly important. However, neither the widespread use of online data collection nor the 2018 overhaul has been accompanied by a break in the series: single-parent families continue to grow at a steady pace while the number of reconstituted families remains stable according to the various sources that have been compared.

With regard to multi-residence situations, the new housing sheet makes it possible in principle to identify the situations of children of separated parents: sole custody by one of the parents without multiresidence, equal residence with each parent (known as alternating residence) and main residence with one parent and a minority part of the time with the other (known as shared residence). Alternating residence is a minority modality for children of separated parents, but is growing rapidly. The results on this topic are consistent with those from other sources. On shared residence, there are fewer sources, with less consistency between them, and the new questionnaire leaves parents a large margin of appreciation on how to count their children in this situation. This makes the results difficult to interpret.

The redesign required adjustments to the method described in a previous working paper for identifying same-sex couples in the census. The method, with the adjustments described here, has been implemented to publish initial census results on same-sex couples.