Size doesn't matter: The fallacy of class-differential fertility in social mobility analysis

One of the key methodological and conceptual constraints in social mobility analysis has been how to differentiate between the 'core' pattern connecting social origins and destinations, and the effect of different distributions of origins and destinations across nations and time points. Conventional sociological wisdom has held that class differentials in family size (and occupational change) impose a significant constraint on the marginal distributions in a mobility table. Due to the logic of the random sampling process, middle class fathers with smaller families are under-represented in mobility tables. It is therefore necessary to approach mobility via the randomly sampled social destinations, rather than as a process moving chronologically from origins to destinations. However, a re-working of the Nuffield Mobility Study data-set is used to show that class differentials in fertility have virtually no effect. The explanation of this surprising finding is shown to lie in class size and the actual distribution of large and small families. The finding implies that if occupational change is left as the only process modifying marginal distributions, it deserves greater attention.