Nationality and origins in French statistics: the ambiguities of categories

The categories used in the social sciences for the description and analysis of the phenomena linked to 'immigration' have undergone far-reaching transformations in recent years. After the exclusive use for over a century of a classification based on legally defined nationality, the introduction of the category of 'immigrant' marked a first break from the French statistical tradition. A second, more radical break will occur with the introduction of references to origins, that is to an individual's ancestry, a situation that is now evolving rapidly. These changes are occurring in a context characterized by the mismatch of the systems of vital registration and categorization created to satisfy different and ultimately unscientific imperatives. To understand the gap that has developed between the categorization used in the official statistics and the scientific questions about migration and its long-term consequences for French society, we examine the history of the classification of immigrant populations. This overview points to the influence of the national model on the categories employed in the social sciences. The second part of the article explores the strengths and weaknesses of the various categories used in the quantitative studies of social phenomena that involve 'immigrants' or people of 'immigrant origin'. The problems associated with the construction of categories based on the origins, usually ethnic, of individuals are examined by means of a non-exhaustive catalogue of recent survey applications.