For the first time since the French population census was redesigned in 2004 (replacing the traditional exhaustive census with annual waves on population samples), a new round of the Famille
(Family) survey was conducted in 2011. The new survey, organized by INSEE entitled Famille et logements
(Family and housing), covers a sample of 360,000 individuals in metropolitan France. The LiLi team (INED-Migrinter) has been closely involved in this operation since the questionnaire design stage and is contributing to survey data analysis. The survey data describe the wide range of family configurations and the spatial distribution of families, notably in cases where family members have more than one place of residence. A sample of this kind offers an opportunity for detailed analysis of changes in family behaviours from one generation to the next and within different social groups. It also provides a means to quantify the frequency of atypical or rare family situations – stepfamilies, adopted children, non-cohabiting couples, etc. – which remain undetected in standard survey samples of between 3,000 and 20,000 people, but which are nonetheless socially significant. Last, the sample design offers scope for analyses at both national and regional levels.
After describing an original research tool that brings together research teams and public data producers, we will present a series of initial findings to illustrate the most original aspects of this survey. This is the first survey of its kind produced by French Statistics that (a) identifies same-sex couples; (b) gives exact details of the conjugal tie (cohabiting union, civil partnership, marriage) and (c) identifies non-cohabiting unions. We will therefore begin by examining the different types of union before focusing on family living arrangements in spatial terms (living in the household most of the time, sometimes, or living elsewhere) to identify national configurations and their geographical imprint.