522.7
Studying Social Stratification Among Retirees. Value and Limits of French Administrative Data.

Friday, 20 July 2018
Location: 716B (MTCC SOUTH BUILDING)
Distributed Paper
Paul HOBEIKA, Université Paris 8, France
This presentation focuses on the interest, from a sociological point of vue, of studying retirement inequalities using administrative data, and more precisely the « Echantillon Interregime des Retraités » (EIR). These data, gathered among the various French pension funds by a government statistical service, allow us to describe and analyse precisely the inequalities among retirees. Compared to social surveys, the EIR has exact values of the various pensions retirees receive. It is thus possible to distinguish between, for instance, one part of the pension received in return of the financial contributions made during the working-life, and another part received as a widower right – pensions who go at 80% to women, and represent 23% of their total pension, 1% for men. Therefore, the EIR is useful for measuring gender inequalities among retirees and decomposing these inequalities between different kinds of pensions.

We will argue that the administrative nature of the data also limits of their value for social stratification analysis. The main issue is the instability of administrative categories, who vary according to the French retirement policies. And whereas sex and age are two variables widely used to describe social stratification among retirees, the classic variables used to describe social status in France do not appear in the literature (profession, skill level, education level). Though assigning a profession to a retired person can seem arbitrary, the EIR, matched with another administrative data set (Insee’s « Panel tous salariés »), can be used to achieve this goal, and thus allowing a study of the inequalities between retirees freed from the unstable categories of the administration.