Social mobility, a phrase that mixes two things

The standard tool cross-tabulates a person's social position with that of their father, sometimes their mother, in a table each cell of which gives a frequency. The French surveys on training and occupational qualification, conducted by Insee, the national statistics institute, since 1964, supply these data.

Those tables allow observed mobility to be measured, and it is substantial in volume: a majority of people occupy a position different from their parents'. That finding nonetheless depends on the number of categories used, since a finer classification mechanically reveals more movement.

Part of that mobility is produced by the transformation of the job structure. When the number of agricultural posts falls and service posts rise, movements necessarily occur.

Social fluidity measures what remains once that constraint is removed: the relative chances of reaching a position by origin. It is what tells us about openness, and it changes far more slowly than observed mobility.

That distinction was formalised by comparative sociology, notably in the book by Robert Erikson and John Goldthorpe published in 1992. It rests on odds ratios, which compare, for example, the probabilities of becoming a manager or professional for the child of a manager and for the child of a manual worker.

French work finds relatively stable fluidity over several decades, with modest variations, while observed mobility has varied sharply with transformations in employment.

That contrast explains why opposing readings coexist on the same subject, depending on the indicator used, without either being inaccurate. Saying that society has opened up and saying that it remains frozen can thus each rest on an accurate series, provided one specifies which.

Older tables compared men's positions with their fathers', leaving women out of scope. Recent analyses include both parents and both sexes, which changes the results.

That methodological development complicates comparison with older work, a difficulty flagged by the authors and rarely mentioned in coverage. It also raises a substantive question: should a household's position be read from a single person, or should those of both parents be combined.

Economic work measures mobility through the correlation between parents' and children's incomes, a finer indicator than categories but requiring matched data across two generations. Those data, long scarce, have become more accessible as tax records have been opened to research.

International comparisons find substantial differences between countries, associated with the level of income inequality, a robust association whose causal interpretation remains debated.